c/EPA
           United States
           Environmental Protection
           Agency
           Office of Research and
           Development
           Washington DC 20460
EPA/600/4-90/017
July 1990
Direct/Delayed Response
Project: Laboratory
Operations and
Quality Assurance
Report for Preparation of
Soils from the
Mid-Appalachian
Region of the
United States

-------
                                                    EPA/600/4-90/017
                                                    July 1990
         Direct/Delayed  Response Project:
Laboratory Operations  and Quality Assurance
    Report for Preparation of  Soils from the
Mid-Appalachian Region of the United  States
                             by
               M.L. Papp  and  R.D. Van Remortel
                        A Contribution to the
                 National Acid Precipitation Assessment Program
            OILS-- KwJroni»ntal Protection Agency
            ^S^on 5,  Library (cPL-16)
            5-SQ' C_ Scat-barn Street, Room
                   1^"  6Q6G4
                              U.S. Environmental Protection Agency
                              Ofiice of Research and Development
                                 Washington, DC 20460
                    Environmental Monitoring Systems Laboratory, Las Vegas, Nevada 89193
                       Environmental Research Laboratory, Corvallis, Oregon 97333

-------
                                        Notice
     The information in this document has been funded wholly or in part by the U.S. Environmental
Protection Agency under Contract Number 68-03-3249 to Lockheed Engineering & Sciences Company.
Additional cooperation has been provided under Contract Number 68-03-3246 to NSI Technology
Services Corporation. It has been subject to the Agency's peer and administrative review and it has
been approved for publication as an EPA document.

     Mention of corporation names, trade names, or commercial pioducts does not constitute
endorsement or recommendation for use.

     This document  is one volume of a set which fully describes the Direct/Delayed Response
Project, Mid-Appalachian Soil Survey.  The complete document set includes the  major data report,
quality assurance plan, laboratory analysis handbook, field and preparation laboratory operations
and quality assurance reports, and analytical laboratory quality assurance report. Similar sets are
being produced for each Aquatic Effects Research Program component  project.  Colored covers,
artwork, and the use of the project name in the document title serve to identify each companion
document set.

     The authors of this document  are affiliated with  the Lockheed  Engineering & Sciences
Company, Las  Vegas, Nevada.  The correct citation of the document is:

Papp, M. L., and R. D. Van Remortel. 1990. Direct/Delayed Response Project: Laboratory Operations
     and Quality Assurance Report for Preparation of Soils from the Mid-Appalachian  Region of the
     United States.  EPA/600/x-xx/xxx U.S. Environmental Protection Agency, Las Vegas, Nevada.
     183 pp.

-------
                                      Abstract
     The  Direct/Delayed Response Project Soil Survey includes the mapping, characterization,
sampling,  preparation, and analysis  of soils  in order to  assess watershed response to acidic
deposition within various regions of the United States. Soil samples from  the Mid-Appalachian
region  were transported to a preparation laboratory for processing before delivery to analytical
laboratories.  This document summarizes procedural, operational and data quality compliance with
the protocols used at the preparation laboratory.  Deviations from  the protocols  and difficulties
encountered are identified and discussed.  Estimates of  uncertainty in the laboratory data are
made,  and the  relation of measurement uncertainty to overall  uncertainty in the  routine  data  is
assessed.  Recommendations are made for program improvement.

     A review of the soils data  suggests that the integrity of  the soil samples was maintained
during  the preparation activities.  In most cases, laboratory personnel adhered to the protocols.
Generally,  measurement uncertainty is a very small part of the overall data uncertainty.

     This  report was  submitted  in partial fulfillment of contract number 68-03-3249 by Lockheed
Engineering  & Sciences Company  under the  sponsorship of the U.S. Environmental Protection
Agency. The report covers a period from September, 1988 to April, 1989.
                                            in

-------
                                     Contents
Section
Notice  [[[   ii
Abstract ...................... .  ........................................  iii
Figures ............................................... . ........ . . , .....  vi
Tables  ............................. ,  ............ , .......... . ........... vi
Acknowledgments ................... . ....................................  vii
Abbreviations and Acronymns ..............................................  viii

1   Introduction  .................................................. . ......   1

     Mid-Appalachian Soil Survey ............................................   2
     Laboratory Objectives .................................................   4
     Organization of the Report  .............................................   4

2.   Preparation Laboratory Operations  ........................................   5

     Laboratory Design  [[[   5
     Personnel .  .................  . .......................................   5
     Equipment Disbursement ................. . ............... . ...... ......   5
     Sample Receipt  [[[   7
     Sample Storage [[[   9
     Sample Drying .................................... , ................ .   9
     Determination of Field-Moist pH (PH_MP)  ................................ . .   9
     Determination of Organic Matter by Loss on Ignition (OM_LOI)  ..... ..... . .......  10
     Determination of Air-Dry Moisture (MOIST_P) ..................  . ....... . .....  10
     Sample Disaggregation and Sieving  ......................................  10
     Determination of Rock Fragment Content (RFJMG and RF_FG)  ...... . ............  11
     Sample Homogenization and Subsampling  .................................  11
     Determination of Bulk Density .................... . .................... . .  11
      Ciod  Method (BD_CLD) ..............................................  11
      Known Volume Methods (BD_KV)  ................. . .....................  12
     Sample Shipment  ...........................................  . ........  12
     Sample Archiving  [[[  13
     Record Keeping  [[[  13
     Data Entry  [[[  13
     Sample Tracking   [[[  13
     Progress Reporting [[[  14


-------
                                             (continued)

Sect/on                                                                           Page

       Special Studies	,	 16
     Data Quality Assessment	,	 17
       Quality Attributes .	 17
       Measurement Quality Objectives	 17
       Uncertainty Estimation for Laboratory Analyses	  . ,	 18
       Measurer-tent Quality Samples	 21
       Technical Systems Audits		  23
       Data Verification	 , .	  24

4.   Results arid Discussion	 29

     Evaluation of Quality Attributes	  29
       Precision	  29
       Accuracy	  32
       Representativesos	  32
       Completeness	  32
       Comparability .  ,		  33
     Assessment  of Data Uncertainty Components		  33
       Additive Components of Uncertainty	  33
       Effect  of Measurement  System on Data Uncertainty	  34

5.   Conclusions and Recommendations	  36

     Laboratory Operations	  36
       Laboratory Design	  36
       Equipment  and Supplies	  36
       Sample Receipt	  36
       Sample Drying ,  , . ,	  36
       Field-Moist pH	,	  37
       Organic Matter by Loss on Ignition	  37
       Disaggregation and Sieving	  37
       Bulk Density	  37
       Sample Batching	  37
       Sample Archiving	,	  37
     Quality Assurance  Program	,	  38
       Training	  38
       Measurement Quality Objectives	  38
       Technical Systems Audits	  38
       Data Verification  .	  38

References	  39

Appendices
    A.  Preparation Laboratory Standard Operating Procedures	  41
    B.  Preparation Laboi-itojy  Data Forms  	  99
    C.  Preparation Laboratory Computer Entry/Tracking/Verincation Procedures ........   110
    D.  Comparison of Soil Sample Hornogenization Techniques	   118
    E.  Determination of Dust Constituents at the Preparation Laboratory Facility	   137
    F.  Kraft Paper and Nylon Mesh Analysis	   143
    G.  Technical Systems Audit Reports	   152
    H.  Quality Assurance Reanalysis Templates	   173
    I.  Laboratory  Quality Control Charts  	   177

-------
                                      Figures


Number                                                                          Page

 1     Layout of the Mid-Appalachian Soil Survey Preparation Laboratory	6
                                       Tables


Number                                                                          Page

 1-1   Analytical Parameters Measured at the Preparation Laboratory	8

 2-1   Contents of Initial Packages of Equipment and Consumable
        Supplies  Sent to Sampling Crews	7
 2-2   List of Equipment and Consumable Supplies Used at the
        Preparation Laboratory	8
 2-3   Particle Density Values for Selected Parent Materials  	 12

 3-1   Preparation Laboratory Within-Run Measurement Quality Objectives	19
 3-2   Primary Horizon Types for Sampling Class/Horizon Groups  	21
 3-3   Samples Placed in Mineral and Organic Soil Batches	22
 3-4   Summary of Achievements for Acceptance Criteria of PH_MP	26
 3-5   Summary of Achievements for Acceptance Criteria of OM_LOI   	27
 3-6   Summary of Achievements for Acceptance Criteria of MOIST_P   	28

 4-1   Imprecision estimates for PH_MP	29
 4-2   Imprecision estimates for OMJ.OI	30
 4-3   Imprecision estimates for MOIST_P	30
 4-4   Imprecision estimates for RF_FG and RF_MG	30
 4-5   Imprecision estimates for BD_CLD and BD_KV	31
 4-6   Multiple Survey Comparison of Imprecision for Rock Fragments
        and Clod Bulk Density  	33
 4-7   Comparison of Measurement Uncertainty Components  	34
 4-8   Delta Values and Ratios for Assessment of Data Uncertainty Components 	34
                                          VI

-------
                                Ackno wledgments
       External peer reviews by  the following individuals are gratefully  acknowledged:   L K.
Fenstermaker, Environmental Research Center,  University of Nevada, Las Vegas, Nevada; and
K. A. Dwire,  NSI Technology Services Corporation, Corvallis, Oregon.

       The authors wish to acknowledge the following individuals for their technical assistance
during the operations and data assessment phases of the preparation laboratory activities: J. J.
Lee, U.S. Environmental Protection Agency, Corvallis, Oregon;  D. G. Easterly and S. C. Black, U.S.
Environmental Protection Agency, Las Vegas, Nevada;  D. S. Coffey and J. S. Kern, NSI Technology
Services  Corporation, Corvallis, Oregon;  M. H. Bartling, Martin Marietta Corporation, Oak Ridge,
Tennessee;  J. V. Burton,  Woodward & Clyde Consultants, Wayne, New Jersey;  and G. E. Byers,
J. E. Teberg, R. L. Tidwell, K. C. Kohorst, G. D. Merritt, G. E. Satterwhite, D. A. Jadhon, J. R. Parolini,
D. L Peres,  M. E. Silverstein,  S. O. Offemson,  A. D.  Tansey,  B. A. Schumacher, and M. J. Miah,
Lockheed Engineering & Sciences Company, Las Vegas, Nevada. We appreciate contributions to
the project overview of this report by the DDRP Technical Director, M. R. Church, U.S. Environmental
Protection Agency,  Corvallis, Oregon.

       The  following  individuals  provided  editorial  and graphics  support and  are   gratefully
acknowledged: B. N. Cordova, A. M. Tippett, K. M. Peres, and  L. A. Stanley, Lockheed Engineering
& Sciences Company, Las Vegas, Nevada.

       Finally, we appreciate the support of our technical monitors, Louis J. Blume and Daniel T.
Heggem,  during the course of this survey.
                                           VII

-------
                      Acronymns and Abbreviations
AERP
BD_CLD
BD_KV
DDRP
EMSL-LV
EPA
ERL-C
FAL
FAO
FAP
kg
LAL
LAO
LAP
MASS
MOIST_P
MQO
NAPAP
NIST
OM_LOI
PH_MP
PLSOP
PVC
QA
QART
QC
QCAS
QCCS
QE
RF_FG
RF_MG
RSD
SCS
SD
set ID
S/H
USDA
Aquatic Effects Research Program
bulk density using clod samples
bulk density using known volume samples
Direct/Delayed Response Project
Environmental Monitoring Systems Laboratory - Las Vegas, Nevada
U.S. Environmental Protection Agency
Environmental Research Laboratory - Corvallis, Oregon
low-range field audit sample
organic field audit sample
mineral field audit sample
kilogram
low-range laboratory audit sample
organic laboratory audit sample
mineral laboratory audit sample
Mid-Appalachian Soil Survey
air-dry moisture
measurement quality objective
National Acid Precipitation Assessment Program
National Institute of Standards and Technology
organic matter by loss on ignition
field-moist pH in deionized water
preparation laboratory standard operating procedures
polyvinyl chloride
quality assurance
quality assurance reanalysis template
quality control
quality control audit sample
quality control check sample
quality evaluation
fine gravel rock fragments
medium gravel rock fragments
relative standard  deviation
Soil Conservation Service
standard deviation
set identification  code
sampling class/horizon group
United States Department of Agriculture
                                          VIII

-------
                                      Section  1

                                    Introduction
     The  Direct/Delayed  Response  Project
(DDRP)  is  an  integral  part  of the Aquatic
Effects Research Program (AERP) of the U.S.
Environmental Protection Agency (EPA).  The
DDRP  is  administered  under  the  federally-
mandated National Acid Precipitation Assess-
ment  Program  (NAPAP) and  addresses con-
cerns relating  to potential  acidification of
surface  waters by  atmospheric  deposition
within the eastern United States.  Given that
acidification  of  some  surface  waters  has
occurred, critical scientific and policy questions
focus on whether acidification is continuing in
the regions  of  concern, whether it  is  just
beginning in other regions, how extensive the
effects might become,  and  over  what time
scales effects might occur.   The DDRP was
designed as the terrestrial complement to the
aquatic resources program and includes soil
surveys encompassing  the  Northeastern  re-
gion, the Southern Blue Ridge  Province, and
the Mid-Appalachian  region  of  the   United
States (Church et al., 1989).  The project draws
its name from  the consideration of whether
acidification might be immediate, i.e., "direct",
or would  lag in time, i.e., "delayed", because of
soil buffering characteristics.

     The overall  purpose of  the DDRP is to
characterize three geographic regions  of the
United States  by predicting the  long-term
response of watersheds and surface waters to
acidic deposition.  The four specific goals of
the DDRP are to: (1) characterize the variability
of soil and watershed attributes across these
regions, (2) determine which  soil and  water-
shed characteristics are most strongly related
to surface-water  chemistry,  (3) estimate the
relative importance of key watershed proces-
ses  in controlling surface water chemistry
across the regions of concern, and (4) classify
the sample of watersheds with regard to their
response characteristics and extrapolate the
results from the sample of watersheds to the
regions of concern.

     The DDRP watersheds were selected as
a subset of lake and stream systems surveyed
in the Eastern Lakes  Survey  and  National
Stream Survey  (Church, 1989) and were char-
acterized as probability samples to ensure that
results could be  extrapolated to a specified
target  population.  A series of maps of soils,
vegetation, land use, and depth to bedrock
were prepared  for each DDRP  watershed by
the United States Department  of Agriculture
(USDA) Soil Conservation  Service (SCS).  Soil
sampling classes were defined for each DDRP
region, and soils selected  from  these classes
were sampled and analyzed for physical and
chemical characteristics  (Lee  et al.,  1989).
Soils were aggregated within sampling classes
to develop class means and variances used in
representing, e.g., by mass or area weighting,
characteristics of the watersheds of interest.

     A variety  of complementary data ana-
lyses  are being  performed as part of the
DDRP,  including the  statistical evaluation of
interrelationships  among atmospheric deposi-
tion, mapped watershed characteristics, soil
chemistry, and  current  surface  water  chem-
istry.  The principal goal of these analyses is
to verify that the processes and relationships
incorporated  in  the   subsequent  modeling
analyses reasonably  represent  the systems
under study.

     Recent  trend  analyses have  indicated
that the  rate of  sulfur deposition  is  slowly
declining in the Northeastern United States but
is increasing in the Southeastern United States
(Church et al.,  1989).   If  a "direct"  response
exists between  sulfur deposition and surface
water alkalinity, the extent of current effects
on surface water probably would not change

-------
much at current levels  of  deposition, and
conditions  would improve  as the  levels  of
deposition decline. If surface water chemistry
changes  in a "delayed"  manner,  e.g., due  to
slow changes in watershed  chemistry,  future
changes  (even with current or declining rates
of deposition) become difficult to predict. This
range of  potential effects has implications for
public policy decisions  on  sulfur emissions
control strategies.

     The three  DDRP  surveys  focused on
regions of the United States that were iden-
tified as potentially sensitive to surface  water
acidification.  The  Northeastern Soil Survey
was conducted in 1985 in the states  of Maine,
New  Hampshire,  Vermont,  Massachusetts,
Connecticut,  Rhode  Island,  New York, and
portions  of Pennsylvania.  The Southern Blue
Ridge Province Soil  Survey was conducted in
1986 in portions of Virginia, Tennessee,  North
Carolina,  South Carolina, and  Georgia.  The
Mid-Appalachian   Soil Survey  (MASS)  was
conducted in 1988 in portions of Pennsylvania,
Virginia, and West Virginia.

Mid-Appalachian  Soil  Survey

     The MASS focused  on watersheds  within
the  states of Pennsylvania, Virginia, and West
Virginia.  The streams  in  this region were
sampled  as part of the  National Stream Sur-
vey, Phase I, which is  an  EPA program de-
signed to evaluate stream chemistry in regions
of the United States believed to be susceptible
to the effects of acidic deposition. A sampling
design was  applied to  allow for  unbiased
characterization of  Mid-Appalachian regional
populations, which resulted in the selection of
36 watersheds for detailed study.

     Soil types that were identified in the Mid-
Appalachian sampling region were combined
into sampling classes which are either known
to have or are believed to have similar chemi-
cal  and physical  characteristics.  Each of the
sampling classes was then sampled across a
number of different watersheds in  which the
classes occur.  Using this approach, a given
soil sample does  not just represent the spec-
ific  watershed from which  it came.  Rather,
each sample contributes to a set of samples
that collectively represent a specific sampling
class for all DDRP watersheds  within the
sampling region (Lee et  al., 1989).
     The  DDRP technical director at ERL-C
had overall responsibility for the MASS.  There
were five sampling crews, each consisting of
three  to four soil  scientists, involved in the
sampling  phase.   In each  of the  150  soil
pedons selected for sampling, all soil horizons
were described  on a coded field data form.  A
total of 844  soil horizons from these pedons
were sampled in accordance with the sampling
protocols. In addition to collecting 5-kilogram
routine  soil samples, each  sampling crew
collected one duplicate  sample for  QA pur-
poses from  every third pedon sampled.  The
sampling protocols and a detailed account of
the soil  mapping and sampling activities are
contained  in a  separate field operations  and
QA report  (Kern and Lee, in press).

     As  part  of the MASS,  a  preparation
laboratory was established to process  soil
samples collected  from  the Mid-Appalachian
region and to perform preliminary analyses on
these samples.  The laboratory was located at
the Lockheed Engineering & Sciences Company
warehouse facility in close  proximity to the
EPA Environmental Monitoring Systems Labor-
atory  in  Las Vegas, Nevada (EMSL-LV) to
ensure the rapid turnaround of samples and to
promote real-time data quality evaluation.  The
preparation laboratory was designed to be the
link between the  sampling crews and  the
analytical  laboratories, and served as a  dis-
tribution  point  for equipment  and  supplies
used by the  sampling crews.

     All participants  were required to comply
with specified  protocols  documented in the
Preparation  Laboratory  Standard Operating
Procedures (PLSOP) manual (see Appendix A).
The laboratory was supervised by a laboratory
manager, who was an experienced soil scien-
tist with  an advanced  university degree  in
soils.    The  manager was responsible  for
ensuring that sample integrity was maintained
after the bulk field samples were  received by
the preparation laboratory.   The  laboratory
analysts  had previous laboratory  experience
and most held a university degree  in the biolo-
gical or  earth sciences.  The  laboratory staff
was responsible  for tracking  soil samples
throughout   processing,  performing   soil
analyses,  preparing samples for shipment to
analytical  laboratories, and documenting all
preparation activities.

-------
      The four types of soil samples sent to
the  preparation  laboratory during the field
sampling phase of the MASS were:
   •  Bulk samples,  those soil samples col-
      lected from specific soil  horizons for
      each sampling  class.  Both mineral and
      organic samples were collected; organic
      soils were defined as having 20 percent
      or more organic matter as determined by
      loss on ignition.  A bulk  sample con-
      tained approximately four  liters of soil
      (eight   liters   for  organic soils)   and
      generally  weighed  about  5  kilograms.
      The bulk samples were collected in large
      plastic bags and placed within protective
      outer canvas bags.   Each  bulk sample
      arrived  at  the  laboratory with a  com-
      pleted DDRP Label A (see Appendix B)
      affixed to the inner plastic bag.
   •  Field audit samples  were measurement
      quality samples sent by the EMSL-LV QA
      staff to the sampling crews to  include
      with the soils sampled from  every third
      pedon by each  crew.  The samples were
      sieved, packaged, labeled,  and shipped
      in the same manner  as their associated
      routine bulk samples.
   •  Clod samples were fist-sized, structurally
      intact bulk density samples collected in
      triplicate,  where  possible,  from desig-
      nated horizons. Each clod was wrapped
      in a hairnet and dipped briefly in  a 1:5
      saran:acetone solution to help maintain
      the clod structure and reduce moisture
      loss from the clod during transport and
      storage.  The clod was labeled and then
      covered  by a  small  plastic bag and
      placed in a special clod box with dividers
      and cushioned  packing material.
   •  Known  volume  bulk density samples
      were collected  in duplicate, where pos-
      sible, from horizons where clod samples
      were  unobtainable.   There  were two
      types of  known  volume samples  col-
      lected in the MASS:  volume-replacement
      and volume-fill samples.  These  bulk
      density samples were packaged in small
      pre-labeled  plastic  bags  in  much the
      same manner as the bulk soil samples.

      The laboratory  staff performed several
soil analyses on the samples, including deter-
minations of field-moist pH in deionized water,
bulk density by clod and known volume meth-
ods, and gravimetric measurements of percent
air-dry moisture, percent rock fragments of fine
and medium gravel size, and percent organic
matter by loss on ignition.  After the samples
were processed and analyzed, homogenization
and subsampling occurred and the  resulting
analytical samples were assembled into batch-
es and shipped via overnight courier to desig-
nated    analytical laboratories   for further
analysis.

     An important function of batch formation
was to place measurement quality samples in
the batches without revealing their identity to
the analytical laboratories.  The measurement
quality samples were  intended to provide a
way to both control and assess data quality
within each batch and between batches.  Both
mineral   and  organic  soil  batches   were
assembled,  each containing a field duplicate
sample,  two preparation duplicate samples,
seven   natural   soil   audit  samples,   and
approximately thirty routine  samples.   The
analytical samples were organized according
to their  respective pedon/set  codes and then
randomized  within each batch by the labora-
tory manager. The samples were labeled only
with batch  and  sample numbers,  and the
batches were subsequently distributed among
three analytical laboratories contracted by EPA
to perform  physical and  chemical analyses
(Byers et al., in press).

     All  information  pertaining  to   sample
receipt,   tracking, preparation,  and  analysis
activities was placed on raw  data forms and
entered  into a computer entry and verification
program.  A personal computer was used to
track samples, evaluate laboratory progress,
calculate final data values, and identify dis-
crepancies in the data.

     An  intensive  quality assurance  (QA)
program  was used at the preparation labora-
tory to maintain consistency in the application
of soil preparation  protocols and to ensure
that the soil  sample  analyses  would  yield
results of known quality. Laboratory personnel
received hands-on training in the protocols and
analytical methods.   In addition,  representa-
tives of the QA staffs at EMSL-LV and the EPA
Environmental   Research   Laboratory   at
Corvallis, Oregon (ERL-C) conducted on-site
technical systems audits of  the  preparation
laboratory.  Daily communication between the
QA  staff  and  laboratory  personnel  was
established  to identify, discuss, and  resolve
issues. Control charts were also developed by

-------
the QA staff to track the quality of data gen-
erated at a given point in time.
Laboratory Objectives

     The primary objectives of the preparation
laboratory were:  (1) to prepare homogeneous,
anonymous,  air-dry  subsamples from  pro-
cessed bulk  samples  received from the sam-
pling crews,  (2) to organize batches of sub-
samples, including quality evaluation (QE) and
quality  control  (QC)  sampler,   suitable  for
shipment to analytical laboratories, and  (3) to
perform  preliminary anaiysas of the bulk  roil
samples to supplement the data generated by
the analytical laboratories.

      This report describes the operations and
data quality  facets  of the  MASS preparation
laboratory program. A series of recommenda-
tions   for  improving  both  the   quality and
efficiency of  preparation laboratories in future
soil surveys is also included. Specific informa-
tion  concerning  the  protocols used  for  the
MASS  sample  preparation can  be found in
Appendix A.

Organization  of the  Report

     This document has been organized into
five  main sections.  Section  1  provides an
                              overview of the survsy objectives and prepara-
                              tion laboratory analytical parameters.  Section
                              2  systematically  describes  the MASS prep-
                              aration laboratory operations.   Section  3 de-
                              scribes tue implementation of  the  overall QA
                              program,  its relation  to data quality assess-
                              ment,  and  the use  of  measurement quality
                              samples  during  the  various stages of  data
                              collection and verification.  Section 4 provides
                              the data quality results and discussion for the
                              analytical parameters measured at the prepar-
                              ation laooratory.  Section 5 provides the con-
                              clusions and recommendations that have been
                              generated from these findings, particularly in
                              ragard to  issues of concern for  sample prepar-
                              ation, improvement in QA design, and sugges-
                              tions  for  QA  efforts   in  future  surveys.
                              Throughout   the  report,  data   quality  is
                              discussed in  terms  of  precision,  accuracy,
                              representativeness, completeness, and  com-
                              parability.  Data  quality achievements in rela-
                              tion to objectives established at the beginning
                              of the survey tsre evaluated.

                                    The chemical and physical soil parame-
                              ters analyzed at the preparation laboratory are
                              referenced either by a data-variable or descrip-
                              tive parameter name.  A list of parameters and
                              their corresponding  descriptions  is  given in
                              Table  1-1.  The  order of  the  parameters is
                              consistent with their  order of presentation in
                              this report.
Table 1-1. Analytical Parameters Measured at ths Preparation Laboratory
Parameter
Description of Parameter
PH_MP      Field-moist pH is determined in a deionized water extract using a 1:1 mineral soil to solution ratio or 1:4
           organic soil to solution ratio, measured with a pH meter and combination eSectrodt,.

OM_LOI     Percent organic matter is determined by loss on ignition and is used to distinguish organic soils from
           mineral soils for analytical purposes.

MOIST_P    Air-dry soil moisture measured prior to disaggregation/sieving at the preparation laboratory is expressed
           as a percentage on an oven-dry weight basis; minaral soils are dried at 105°C, organic soils at 60°C.

RF_FG      Fine gravel is the portion of the rock fragments in the soil with particle diameter between 2 mm and 4.75
           mm, and is measured gravimetrically.

RF_MG      Medium gravel is the portion of the rock fragments in the soil with particle diameter between 4.75 mm and
           20 mm, and is measured gravimetrically.

BD_CLD     Bulk density is measured from replicate intact clod samples collected in the field, and  is measured by a
           gravimetric water submersion procedure.

BD_KV      Bulk density is gravimetrically measured from replicate known volume samples collected bv cithsr a volume-
           fill or volume-replacement procedure.

-------
                                     Section 2

                    Preparation Laboratory Operations
     Brief explanations of the specified meth-
ods  and procedures  for the soil preparation
tasks are  given  below in  conjunction with
descriptions of the operational components.
The  procedures are discussed  in sequential
order of performance by the preparation labor-
atory.  The detailed  methods and analytical
procedures used  in soil preparation  for the
MASS are provided in the PLSOP (see Appen-
dix A).   This  section describes preparation
laboratory operations, including difficulties with
the procedures and deviations from the proto-
cols.

Laboratory Design

     The MASS soil processing and analyses
were performed in a secure, access-controlled
warehouse located in  Las  Vegas,  Nevada.
Sample  integrity was ensured by the use of
detailed sample labels,  by  documenting the
status of each sample  during the processing
and  analyses, and by avoiding physical or
chemical contamination during each processing
step.  The warehouse  contained a group of
mobile  laboratory units that were used for
specific  preparation activities, as shown in
Figure 1.  Fume hoods and exhaust fans were
provided in each unit and were operable during
stages of sample processing as required by
the PLSOP.   Ventilated sample drying  and
sieving rooms  were  constructed to aid in
preserving sample integrity.   The rooms were
constructed of wood frames and sheets of
clear plastic, and were equipped  with electric
fans and  filtered  inlet  louvres  to allow air
exchange.  Deionized water and  a source of
filtered compressed  air was available for
routine cleaning of equipment.
Personnel

     The preparation laboratory manager was
required to  be knowledgeable in laboratory
methods and procedures, and have demonstra-
ted ability to track large numbers of samples
and supervise laboratory personnel.  Adequate
staffing was provided to ensure a fast and
efficient turnaround of samples from the field
to the  analytical laboratories.  All personnel
were thoroughly trained  in the protocols and
procedures by the laboratory manager before
sample preparation began.  The laboratory
was able to maintain a steady rate of proces-
sing with a regular  staff  of  four  full-time
analysts and two part-time technicians.

     With few exceptions, the distribution of
tasks followed closely the outline of activities
provided in the PLSOP. The laboratory mana-
ger was  responsible for  assigning duties
according to the specific project needs at any
given time during the operation of the labora-
tory.

Equipment Disbursement

     A laboratory technician, under the direc-
tion of the  laboratory manager and  the QA
logistics  coordinator, was  responsible for
distributing and tracking equipment and sup-
plies to the  sampling crews and preparation
laboratory personnel.  The  equipment cache
was stored in an  area of the warehouse
having  restricted access  and a series of  sign-
out sheets was used.

-------
                              SAMPLE  DRYING
                          NON-DORP
                                                                       OlSAQOnEOATlON
                                                                       SIEVING
                                                                       ROCK FRAGMENT DETERMINATION
                                                             NW  DOOR
                  01-00
                  BULK
                  DENSITY
                                                              LOSS
                                                              ON
                                                              IGNITION
                                                   SAMPLE
                                                   nECEIPT
                                                  NON-DDRP
                                                                     AFP
                                                                            WC
                                                                            we
                                                                I
                          PARKING AREA
                                                                                        HOMOOENIWTION
                                                                                        SUB-SAMTL1NO
                                                                                                    ROLL
                                                                                                    DOOR
                                                                                           NW  DOOR
fi
                                                                     FRONT  OFFICE
                                                                                                   c
Figure  1.  Layout of  the Mid-Appalachian Soil Survey Preparation  Laboratory.

-------
      An initial package consisting of all essen-
tial sampling hardware and a two-week supply
of the necessary consumables  was prepared
and distributed to each  of  the original four
sampling crews at the beginning of the survey.
A second shipment, containing a major portion
of the remaining consumables needed during
the survey, was sent to the appropriate United
States Department of Agriculture (USDA) Soil
Conservation Service (SCS) field  office  or
another convenient delivery point approximately
one week after the first shipment.  Each  initial
field  equipment package included the  items
presented in Table 2-1.
Table 2-1. Contents of Initial Packages of Equipment
          and Consumable Supplies Sent to
          the Sampling Crews
Equipment/Consumables
Quantity
Hand pump                             1
35-mm camera w/built-in flash              1
Clod boxes                            30
Brass sieve, square-holed, 19-mm            1
Stapler, large                            1
Stapler, small                           1
Gel-pacs                              80
Coolers                               10
Hole punch                             1
Straight-edged trowel                     1
Cylinder, clear plastic, 250-mL              1
Graduated cylinder, clear plastic, 2000-mL     1
Macroballoon beads, 3.18-9.53 mm diameter    4 L
Plug for packing beads in cylinder            1
Field pH kit                             1
Random number table                     2
SCS-SOI-232 field data form               50
Orange  flagging rolls                     10
Yellow marker flags                      80
Indelible marker - thick                   10
Indelible marker - thin                    10
Golf tees (horizon delineation)             20
Photogray cards - 8" x 10"                 20
Horizon depth tape                       1
Metal paint can, teflon-lined, w/lid, 3.8-L       2
Hairnets                              144
Plastic inner sampling bags               350
Canvas soil bags                       350
DDRP Label A                         200
Staples, box                             2
Small plastic clod bags                  600
Twist ties                             300
Clod tags                             300
Clod box labels                        500
Log books                              5
Saran™ powder, 540-g packets             10
Acetone, 19-L can                         1
     A fifth sampling crew was brought into
service later in the sampling period.  This did
not  cause  any serious difficulties with  the
distribution of  equipment  and  supplies, al-
though two of  the  other four crews had to
share  some of  their supplies and field audit
samples with  the newest crew.  The sampling
crews usually obtained supplies by calling the
laboratory manager at  least one week before
an anticipated shortage or  by  making these
needs known on the weekly conference call.
On occasion,  a crew member would  request
additional  supplies  when a  particular consu-
mable was already depleted or could not be
found.  All shipments to the crews were made
by an overnight carrier  service.

      Supply  sheets were  provided  for  the
sampling crews, who were requested to log all
consumables  taken from or returned  to  the
SCS offices, which allowed better estimates to
be made of any additional consumable items
needed.  Log entries identified the sampling
crew, the equipment taken, and the date.

      Preparation  laboratory personnel were
originally tasked  with  providing  the  saran-
acetone mixture used for coating bulk density
clod samples  in the  field. Because of leakage
and  safety problems in the  containment and
transport of the pre-mix solution, it was deci-
ded  that the  laboratory would  instead send
pre-weighed  packets of saran  powder  and
factory-sealed cans  of acetone.  The sampling
crews prepared their own mixtures of approxi-
mately 1:5 by weight of saran:acetone solution.

     After soil  sampling was  completed, all
hardware and remaining consumables were
returned to the preparation  laboratory.  The
hardware and supply items  listed in Table 2-2
were required to perform the prescribed soil
processing and  analysis tasks.

Sample Receipt

     To facilitate the receipt of sample ship-
ments,  the sampling crews were asked to
make every effort to inform the preparation
laboratory  when samples were being shipped
from the field.  Because this was often logisti-
cally difficult,  the preparation  laboratory in-
itiated an alternating work week that enabled
laboratory personnel to  be available to receive
shipments  during all carrier delivery hours, six
days per week.  As a result, no carrier ship-
ments had to  be turned away because of the
laboratory being closed.

-------
Table 2-2. List of Equipment and Consumable
        Supplies Used at the Preparation Laboratory

  Items

Log books
Raw data forms
Munsell color book
Pens, indelible black ink
Kraft paper, 1-meter (36") wide rolls
Plastic gloves (unpowdered)
Dust masks
Organic vapor masks, cartridge-type
Metal paint cans, teflon-lined, w/ lid, 1-gallon
Aluminum weighing tins
Nalgene bottles: 2-L, 500-mL, 125-mL
Carboys,  13-gallon capacity
Crucibles, 450"C tolerance
Evaporating dishes, 450°C tolerance
Beakers,  1000-mL
Tongs
Ring stand
Thermometers, 0 to 100° C range
Sample drying tables, PVC / nylon mesh
Rolling pins, wooden
Rubber stoppers, No. 10 size
Brass sieves, square-holed: 2-mm and 4.75-mm
Riffle splitter, closed-bin, with 1.25-cm baffles
Top-loading balance, 0.01-g accuracy, 500-g capacity
Top-loading balance, 0.1-g accuracy, 5-kg capacity
pH meter, with proper electrodes
Desiccators
Desiccant
Hot plate
Convection oven
Muffle  furnace
Shipping  boxes
Packing material
Strapping tape
Saran™ powder
Acetone
Plastic containers, 25-mL and 50-mL
      Bulk  soil samples,  known volume  bulk
density samples, and field audit samples were
placed  with  two  frozen gel-pacs  in  box-
enclosed styrofoam coolers. Sample shipment
forms and the original SCS-SOI-232 field data
forms (Kern et al., in press) corresponding to
the samples were sealed in plastic bags and
taped to the  underside of the cooler covers.
The outer boxes enveloping the coolers were
firmly enclosed with strapping tape and the
coolers were  then  shipped  directly to  the
preparation laboratory via  overnight carrier.
Because of their fragile nature, the  soil  clod
samples were securely packed with vermiculite
or crumpled newspaper  in coolers  separate
from  the bulk soil samples.  Each sample was
labeled in  accordance  with the  protocols
outlined in Appendix A.
     The field data forms were completed by
the sampling crews  while at  the  sampling
sites. Each form listed all available site infor-
mation for a particular pedon,  including des-
criptions of landform, vegetation, soil climate,
and a detailed pedon description.  The sample
labels  were checked  for  coding  accuracy
against the field data  form and sample ship-
ment form.  A sample receipt raw data form
(see Appendix B) was filled out  completely for
all samples  received  and logged in  at cold
storage.

     All   preparation   laboratory  personnel
became  familiar  with the  sample  labeling
system,  as it  was the  primary  information
used in receiving and tracking samples. After
receiving a sample shipment, the  preparation
laboratory  staff   checked  the  condition  of
incoming  samples and  logged the  sample
labeling information on a sample receipt form.
The  number of samples collected  and the
horizon, watershed, and sampling class desig-
nations were compared  from  the field data
form to the sample labels to ensure that all of
the samples collected had actually arrived with
the shipment and that the  labeling was cor-
rect.  The field data forms  and sample ship-
ment forms were manually scanned for consis-
tency and any discrepancies were noted and
brought to the attention of the QA manager,
although a copy of each form was kept at the
preparation laboratory.   Raw data from the
sample receipt form were entered into  a data
entry and verification program on a  personal
computer  (described  later in  this  section).
Discrepancies noted on the verification printout
were evaluated  and discussed with  the  QA
manager and the ERL-C sampling QA staff.

     While the laboratory manager validated
the accuracy of the sample labeling, the sam-
ples were grouped according to their respec-
tive pedon and set identification  (set ID) codes
and  were  placed  in cold storage.   The  QA
manager was notified  of any labeling discrep-
ancies and immediate consultations  with the
ERL-C sampling QA staff or the  appropriate
sampling crews helped to resolve the inconsis-
tencies.   Subsequently,  the samples were
allowed to undergo the next stages of sample
preparation.

-------
Sample Storage
     A walk-in cooler was constructed within
the preparation laboratory in order to contain
both the  incoming  field  samples  and  the
samples being  stored  in the  intermediate
stages of processing. A separate off-site cold
storage facility was used to  store archived
subsamples from those bulk samples that had
already been processed  and batched.

     Upon  receipt,  all  field   samples  were
placed in cold storage as soon as possible
and remained there until sufficient space was
available in the drying  area   for a "run"  of
samples (ranging  from 20 to 40 samples,
grouped according  to set ID) to be air dried as
described below. After air drying, the samples
were returned to  cold  storage  until further
processing.  Samples were organized in cold
storage so that  a  particular sample within a
given set could be easily located.  This was
best accomplished by grouping the samples
according  to their  component set and pedon
codes which were  maintained during storage,
processing, and shipping.

     The cold storage area was maintained at
a constant temperature of 4°C and was  moni-
tored by a continuous rotation sensor and by
daily thermometer  readings.   A substantial
deviation in temperature  (±  2°C  or more)
occurred only once and  was  recorded in the
cold storage log book. Specifically, the labora-
tory had a temporary refrigeration  failure on
April 24, 1989  that was reported to the QA
staff immediately.    The unit was repaired
within  approximately four hours.    Sample
integrity was  not  threatened  because  the
failure occurred after sample  processing had
been completed on the bulk samples.  At the
time, only clod samples were in the facility and
they were  not considered to be temperature-
sensitive.

Sample  Drying

     When  there  was   sufficient space in
the  enclosed   drying area, the  field-moist
bulk samples were taken from cold storage,
grouped by pedon code,  and were spread out
on large sheets of  kraft paper placed on
drying tables to air  dry.  Where necessary,
several layers  of paper were placed under-
neath the samples  to absorb excess moisture
and this paper was  replaced periodically for
faster drying.  A sheet of paper was placed
loosely  over  each sample to  minimize  the
possibility  of cross  contamination  among
samples without restricting air circulation.

     The use of nylon mesh drying tables and
a multiple-fan air exchange system enhanced
the air circulation around the samples, thereby
increasing the rate of sample drying.  Labora-
tory personnel occasionally stirred the samples
to encourage uniform drying. Possible sources
of sample contamination, such as chemicals,
food, drinks, and smoking, were prohibited in
the drying area. A pair of gloves dedicated for
each sample was always worn when handling
the samples.   The drying  area  was damp-
mopped  at least once  per week to  control
dust, and was done after air-dry samples were
returned to cold storage and before field-moist
samples were laid  out to dry.

     Organic  soils in advanced stages  of
decomposition and mineral soils high in clay
had a tendency to  harden nearly irreversibly if
allowed to  dry without  concerted efforts  to
disaggregate   clumps  and  structural peds
common in these types of samples.  With this
in  mind, efforts were made  during  the  air-
drying  period to disaggregate these clumps
and peds while the samples were still some-
what moist or friable before reaching an  air-
dry state.

Determination of  Field-Moist pH
(PH_MP)

     Immediately  after  a field-moist  bulk
sample  was  spread  to  air-dry,  a composite
subsample  was collected  from across  the
drying  sample and placed  in a  125-mL Nal-
gene™ bottle.  The pH of the sample in deio-
nized water was determined in a mobile labor-
atory unit set  up specifically for this measure-
ment.   A digital  pH/mV meter, capable  of
measuring pH to 0.01 pH unit using a combina-
tion pH  and reference electrode, was calibra-
ted immediately prior to  each  run of analysis
using  standard stock buffer solutions of pH
7.0 and pH 4.0.  Approximately 1:1 and  1:4 soil
to  solution ratios  were used for  mineral and
organic  soil analysis, respectively.  The elec-
trode  was rinsed  with deionized water  bet-
ween each sample  in order to prevent solution
carryover.

-------
     Early in the survey, the laboratory ana-
lysts experienced difficulties in the calibration
of the meter. The problems were traced to the
group of available electrodes that appeared to
be  faulty  because  of  fluctuating  readings.
While additional electrodes were being requisi-
tioned, a borrowed  electrode was used and
provided  sufficient  precision  and  accuracy
results.

     All calibration readings and adjustments
were  recorded in a  soil pH logbook and ini-
tialed by the analyst.  Data from each day's
analysis were recorded on a pH raw data form
and  subsequently entered  into a data entry
and verification program on a personal com-
puter.

Determination of Organic Matter
by Loss on Ignition  (OM_LOI)

     A loss-on-ignition  method was used to
determine percent organic matter in the sam-
ples and to classify each sample as a mineral
or organic  soil.  Because organic soil con-
stituents are oxidized at  high  temperatures,
percent organic matter could be calculated on
a weight-loss basis. From this, the organic
carbon content can be estimated.  A modified
version of the method described in MacDonald
(1977) was  used. This method  was added to
the DDRP preparation laboratory activities for
the MASS in order to identify organic samples,
containing 20 percent or more organic matter,
that would be grouped  into separate batches
for analysis.

     A composite  subsample of a bulk soil
sample was placed into a crucible and dried
overnight in an oven to determine its air-dry
weight.  Samples that were suspected to be
mineral soils were dried at  105°C and organic
samples were dried at 60°C. After weighing,
the crucibles were  placed  in a 450 °C muffle
furnace activated in an operable fume hood.

Determination of Air-Dry
Moisture  (MOIST_P)

      Each  bulk sample was allowed to dry
until it appeared to be at or below the speci-
fied  air-dry moisture content,  which was 2.5
percent for mineral soils and 6.0 percent for
organic soils. Air drying generally took about
two or three days for most samples, although
certain saturated organic samples took a week
or more to visibly dry.  At this time, a com-
posite subsample was selected for the gravi-
metric determination of air-dry moisture.  If the
moisture  criteria  were not  satisfied  for a
sample, it was allowed to continue drying until
a later determination could be made. In a few
cases, certain samples would not  dry  suffi-
ciently to  satisfy the criteria but were  con-
sidered to be air dry for practical  purposes.
Differences in soil texture and mineralogy are
thought  to have influenced  this  variation in
moisture content.

Sample  Disaggregation and
Sieving

     After a bulk soil sample was determined
to be air dry, it was ready to be disaggregated
and sieved in order to remove rock fragments
and to prepare the sample for homogenization
and subsampling.   The disaggregation and
sieving tables were  covered with  layers  of
kraft  paper  that were replaced after each
sample had been sieved.   Dust  masks and
protective clothing were worn at all times, and
ventilation and heating was  provided for the
sieving area.  All equipment  used during this
operation  was cleaned with compressed air
and fresh tissue  paper after each sample had
been sieved.

     After recording  the weight of the  air-dry
bulk sample, the soil was  disaggregated to
allow  the  less than 4.75-mm soil fraction to
pass  a No.  4 sieve and the less than 2-mm
soil fraction to pass  a No.  10 sieve.  Depend-
ing on the consistence of  rock fragments in
each  sample, a  wooden rolling  pin or large
rubber stopper was used for disaggregation.
The rolling pin was  successfully used  when
there were no fragments in the sample or the
rock fragments were  extremely firm.  Because
of the highly weathered state of  many of the
shaly fragments  in  some  samples, various
types and sizes of rubber stoppers were used
to selectively disaggregate the individual soil
peds. This procedure allowed the passage of
soil through the sieves without breaking the
fragments.
                                           10

-------
Determination of Rock Fragment
Content  (RF_MG and RF_FG)

     Rock  fragments  remaining  from  the
sieving procedure were retained on two sieves;
the No. 4 sieve retained the 4.75- to 20-mm
medium gravel fraction and the No.  10 sieve
retained the 2- to 4.75-mm fine gravel fraction.
Fragments larger than 20 mm were sieved out
of the  samples by the field sampling crews.
Gravimetric analysis  was used to determine
the percentage by weight of each fraction in
the bulk soil samples.

     There were no deviations from the speci-
fied protocols for rock fragment determination.
The laboratory bagged and labeled the frag-
ments  and  awaited  further instructions for
dispensation.

Sample  Homogenization and
Subsampling

     In order to obtain representative quan-
tities of soil suitable for physical and chemical
analysis  at  the  analytical  laboratories,  the
protocols specified the use of  a  closed-bin
riffle splitter  to prepare homogenous subsam-
ples from the less than 2-mm  soil  fraction.
The  riffle splitter  was cleaned with com-
pressed air after each sample was split.  The
compressed air was passed through a desic-
cant filter trap. If the operator suspected that
the riffle splitter was still not thoroughly clean,
the riffle  splitter was washed with deionized
water and allowed to air dry completely.  The
work area was vented with an operating fume
hood and the surfaces were  cleaned with  a
hand brush after each sample was processed.

     The preparation laboratory prepared  two
different  subsamples because separate con-
tracts were awarded to analytical laboratories
for the general analysis of 47 physical  and
chemical parameters and  for the elemental
analysis of total carbon, nitrogen, and sulfur.
The general analytical  sample weighed  ap-
proximately 400g to 500g and the elemental
analytical sample weighed  approximately  30g
to 60g.

     A preparation duplicate sample was split
from each  field duplicate  sample and  also
from its associated routine  sample.   The
preparation duplicates served as measurement
quality samples for the purposes of evaluating
the precision associated with the homogeniza-
tion and subsampling procedures.

     There were no deviations from the speci-
fied  protocols for soil homogenization and
subsampling.   A daily record of the riffle
splitting operation was made in a logbook for
each sample, specifying the initial weight of
sample and a description of the passes made
through the splitter to achieve the final weight
of the analytical samples.

     The  soil  material remaining after sub-
sampling the analytical samples  was placed
into a 2-L Nalgene bottle for archiving in long-
term cold storage.

Determination of  Bulk Density

     For the DDRP surveys, the bulk density
of a  soil is defined as the weight of oven-dry
soil (minus rock  fragments) per  unit volume
including the pore space, expressed as  g/cm3.
For most  mineral soils,  bulk  density  ranges
from about 0.6 to 2.0 g/cm3. With increasing
organic matter content, soils generally exhibit
a decrease  in bulk density because organic
matter has higher porosity and lower density
than mineral particles of the same diameter.

     As in past DDRP surveys, the clod meth-
od was the primary  method for  determining
bulk  density in the  MASS.   Two alternate
methods were also attempted for soil horizons
that failed to yield satisfactory clods. The first
was  a  volume-replacement  method  which
utilized a known volume of small foam  beads
packed into  a  cylinder to replace a selected
volume of  soil excavated from a given horizon,
and the second was a volume-fill method that
involved a simple container filling procedure
which was used if the clod or volume-replace-
ment methods did not produce representative
samples.

Clod Method (BD_CLD)

     Analysis of  the clods was based on the
method described in USDA-SCS (1984).  Repli-
cate  soil clods, usually three per horizon, were
collected by the  sampling  crews whenever
possible (Kern and Lee, in press). The clods
were weighed at the laboratory and then oven-
dried.  Each clod was  dipped in a  1:5  by
weight sararracetorie mixture and was allowed
                                          11

-------
to dry briefly.   The dipping procedure  was
repeated until each clod was assumed to be
impervious to water (usually three to six trips).
Each clod was submerged in a 2000-mL  bea-
ker  of water tared  on a balance and the
weight of the clod was  recorded.  After two
hours in  a muffle furnace at 400°C, the clods
were passed through a 2-mm sieve to detei-
mine the percentage by weight  of  rock frag-
ments  used in calculating the bulk density.

     There  were  no major  deviations from
the specified  protocols  for the  bulk  density
determination.   A number  of floating clods
were identified, mostly when the bulk density
was less than 1.0  g/cm3.   Such clods were
forcibly  submerged  to  allow  bulk  density
determinations to be made.  This was ac-
complished  by fitting the clod into  the lower
cavity of  a test tube holder, such that the clod
would rest against the holder when submerged
and remain  stationary within the beaker.

     Extra care was taken in the use of the
Saran™ powder and acetone solution.  When
mixed,  the resulting solution has a tendency to
volatilize  hydrogen chloride gas which  can
cause deleterious health effects.   The solution
was mixed and  used  only in an  operating
vented  fume  hood,  and  a cartridge-type
organic vapor mask, gloves,  and  laboratory
coat were always worn.

Known  Volume Methods (BDJCV)

     Bulk density was also determined by two
known volume methods whsn clods could not
be collected in the field, as described in  Kern
and Lee  (in press).  The initial and final vol-
umes of  beads used for the volume replace-
ment samples were recorded in the field by the
sampling crews.  The volume filling container
used was a 250-mL graduated plastic beaker,
which had an absolute volume of  300cm3 when
entirely filled.   The  known volume samples
were processed in a manner similar  to the
method described in Blake (1965). The prepar-
ation  laboratory oven-dried and  weighed the
samples, sieved the  rock  fragments,  and
calculated the bulk density.

     A procedural change  was  made in the
calculation  of  the  bulk  density for  known
volume samples.  Instead of using the stan-
dard  2.65 g/cm3  particle density  value as
stated in the PLSOP,  actual rock fragment
estimates grouped by parent mrusra! cia'r;
were used. It waz assumed th3> Ur^o vnr ^.-n
volume  samples  gsnern'ly  v-..-"irj c-'sV.r.l"  a
higher than average per:.,-;?-;.;;--/,-? of ">"-••  •.. M-
ments  and, therefore,  ^ ->rj(' •  • v^;.-* . >o.  •
accurate information o?i ,, ".. :.:-j;. •? T>S''V   '.?>.;
density  was  detenuirMK"    •  •  .',••..•".'-;:•
submersion technique on •• ••,  •;, •;) -<':-)>,,•: --;'. •- •<
from the bulk &,?mp!" r-'   ••..-.:;  ,i^-:'-'-  J;--'
which a known, vo^ur-j ----.I,-. ;, .;-;> r  ,-- ,.'-iv-.ci-"'.
The particle der^:'-5/ ».::, '•"-»,..'•- ~ •••> ;"r-.r.r: •:, :-
differing pys-.,:rt ?r:-3tr -;  r,  --;r;  -• >::!"   :^•';;•_
values YtGi j CCt.V" .'>'U :,.'r  , '  '~ ;•"• .":?*, •'.''.'•  '' -"-.
erial classes,   she  aifif-;a«"t r;itk;i-j t^'isS'ty
vabes  that wore u>t'1  &>'~i Ji--  •?;,<  ''ri r;1-/.^
Table 2-3.  Partlclfc Density VaSu»*
         Mat©rEaE«g
                                      Parent
     Parent Matehal

AO - Sandstone (unspecified)
A1 - Sandstone (noncalcareous)
BO - Tnterbeddod eedimentary
B5 -- Sandstone and shalo
B6 - Sandstone aiid siltstono
B7 - Shale and siltstona
MO - Matarncrphic (unspecified)
M9 - Quartets
                          Particla density (g/cm3)

                                2.52
                                y..$2
                                2.-r4
                                £.44
                                2A4
                                2.45
                                2.51
                                2.51
Sample Shipment

     After the sample  preparation and ana-
lyses were completed for  a sufficient number
of sets of samples, batches were constructed
and  readied  for  shipment  to  the analytical
iaboratories.  A batch of samples consisted of
approximately 40  samples, including about 30
routine samples and 10 different measurement
quality samples (see Section 3) which  were
used to control and assess the various  com-
ponents  of measurement uncertainty.   The
preparation laboratory manager assisted the
QA  logistic?, coordinator  on  packing  and
shipping the batches of samples via overnight
carrier.  The batches of samples were packed
in sealed  heavy-duty  cardboard  boxes for
shipment to the designated recipient analytical
laboratories.   The  QA soil  chemist routinely
called one or two days in advance  to notify an
analytical laboratory that  a b?tch  of samples
was to be  shipped.
                                            12

-------
Sample Archiving
     After  e?inh  batch  of samples  was
shipped, the remaining  volumes of bulk  soil
sample material  were  placed  in long-term
archive. Samples were stored in 2-L Nalgene™
bottles, rffixed  with  a  ODRP  Label  A (see
Appendix  /')  and were grouped in boxes by
their respective batch, set, and  pedon codes.
Multiple bo'-.'Jes from the same sample code
were  -so r.ov,.d en tho iabei.   The boxes of
sample:; '• -c;c Ir,,-:,>-?diate!y placed in Song-term
co'o r/;, . :,  ,.-•., •.';.-'-., oo,u~. warehouse facility
v v.;i ,r'.-••-.,-  -  '   •..,-•-. ;. ir...-:;.^?;1 ?:;? lo a^.-v
     Precision and accuracy of the measure-
ment quality samples were checked while the
samples were  being processed  which per-
mitted the rapid identification of discrepancies
and  allowed the QA  manager  and laboratory
manager to efficiently track progress at the
laboratory.  Features of the  computer entry
arid  verification  program  are highlighted in
Appendix C.

     The  computer program was capable of
generating the thai  batch data  lorn";  ODRP
Form 101 (see Appendix B), whk.n vur>. ;-• fired
fw analytical data hv b;?;tch ;\  ih': c.   .- •.•,.'
                        '
     Seven raw data forms were designed by
EMSL-LV for use  in sample tracking and for
entering data relating  to sample receipt, bulk
sample  processing,  field-moist pH,  air-dry
moisture, loss on ignition, clod bulk density,
and known volume bulk density.  All raw data
were  recorded  ors these  forms in  indelible
black ink and the completed forms were or-
ganized in binders. An example of  each form
is provided in Appendix B.

     The preparation laboratory was provided
with logbooks for recording  raw data on a
daily  basis.   The bulk  sample  processing
forms, plus sampling labels from the original
bulk sample, were placed in binders organized
by batch number.

        Entry

     A  personal computer was used at  the
laboratory for entry and  verification of  all
preparation and analysis data recorded on the
raw data forms.  Entry of the raw data was
accomplished by selected laboratory personnel
who WP?S trained to operate a custom SAS-
AF™ menu-driven program for each  of  the
primary phases of  sample  preparation and
analysis.  The program  provided an efficient
way to track the status of the samples from
the time of field sampling through shipment to
the analytical laboratory.  By comparing data
among  different files,  sample labeling dis-
crepancies were quickly identified and reported
to  the QA manager c;r-d the ERL-C  sampling
QA staff.
manually for ee.uh l>a'>r:h,   >;v- C •'". ••,', ,y  ;r  ,-.
identified the batch as containing mineral  or
organic soil samples and denoted each sample
only by batch and analytical sample numbers.
"this procedure disguised the originating pedon
and horizon of each routine sarnpie as well  as
the identity of the  measurement quality sam-
ples.   The QA manager was provided  with
copies of both forms to serve as documenta-
tion of the shipment.

Sample  Tracking

     As part of the data entry and verification
program,  a  sample  tracking  system  was
implemented in order to document the transfer
of  all  samples  between  the  field  and the
preparation  laboratory and, ultimately,  their
transfer to the analytical laboratories.  Before
sampling began, the DDRP soil sampling task
leader provided the EMSL-LV QA manager and
preparation laboratory manager with  a list of
watershed and pedon  identification numbers
for each pedon to  be sampled.  The informa-
tion was entered into the computer entry and
verification program at the preparation labora-
tory before samples were received.  As the
samples arrived at the preparation laboratory,
the sample receipt form was filled out and
entered into  the  computer.   If  a particular
watershed/pedon combination had been previ-
ously   used,  e.g.,  the preparation laboratory
had previously received  samples  with  that
watershed/pedon  number   or a  particular
watershed/pedon number could not be found
on the ERL-C list, the soil sampling task leader
and the QA manager were notified immediate-
ly.  Early in  the  survey, a series of samples
from nine  pedons were determined  to  have
been labeled with incorrect pedon codes that
                                           13

-------
were different from the codes initially provided.
The appropriate coding changes were made on
the sample labels by the laboratory staff.

Progress  Reporting

     The  preparation  laboratory data  entry
and verification program, described  in Section
3,  produced summary statistics in order keep
the EPA management team  informed on the
progress of field and preparation  laboratory
operations. Upon request, the EMSL-LV techni-
cal monitor was  provided with information on
the total number of sets, pedons, and samples
received by the preparation laboratory, and the
percentage of each laboratory analysis  that
had been completed for the pedons specified
for sampling.  The sampling information  was
used to assess the progress of the sampling
crews. It was necessary to ensure a steady
rate of progress because crews were sampling
during late fall in areas  that would not be
conducive to winter sampling.  The data pro-
ved useful in  showing the need for another
sampling crew to be added  in order to com-
plete the sampling in the allotted period.  The
analytical  information  was used to evaluate
preparation  laboratory progress.   With this
information, the  QA manager and  technical
monitor could determine  where  additional
effort was needed and take appropriate action
to improve the rate of  progress.
                                           14

-------
                                      Section 3

                         Quality Assurance Program
     A preparation  laboratory QA program
was specified  during  the  course of MASS
activities to ensure that data quality could be
evaluated in real time.  The QA program  con-
sisted  of operational and assessment com-
ponents that were meant to aid in producing
data that would  satisfy  the  quality require-
ments  of the DDRP data users. The following
subsections explain aspects of the QA  pro-
gram for preparation laboratory activities.

Operational  Components

     The  operational  components  of  the
preparation laboratory QA  program included
training the laboratory personnel in the proto-
cols to be followed, establishing a communica-
tions  network  between the  laboratory  and
other MASS participants, setting measurement
quality  objectives,  and  performing on-site
systems  audits.  In addition, preventive main-
tenance  and  corrective action were accom-
plished where necessary.

Training

     The PLSOP  (see Appendix A) served as
the protocols for sample  preparation in the
MASS.  Prior to the receipt of routine samples,
all laboratory personnel were initially trained by
the preparation  laboratory manager at a  two-
day hands-on workshop held at the laboratory
on September 9-10, 1988. The purposes of the
workshop were  to familiarize the participants
with the  laboratory  protocols  and  to discuss
key activities and responsibilities.  Training
established a basis for consistency among the
analysts, thereby increasing the likelihood that
the data  would  be of comparable quality.
Additional training sessions were  held every
two weeks at the laboratory, where the labora-
tory manager  clarified the preparation  and
analysis  protocols  and trained the staff  in
other aspects of soil genesis and morphology
that were needed to identify characteristics of
the many  varied soil horizons the staff  was
processing.  Of particular importance was the
identification of weathered rock fragments so
that laboratory personnel would  use  care to
avoid crushing this material during the sieving
process.

Communications

     Additional support was provided through
communications with other DDRP personnel.
The EMSL-LV QA manager periodically visited
the preparation laboratory to discuss the data
quality assessment procedure with the labora-
tory personnel.  The laboratory manager called
the QA manager or other QA representatives
when additional information  was  needed
regarding  a particular  procedure.  The  soil
sampling task leader and QA manager were
notified whenever there  was a problem  con-
cerning the field sampling operation.

     Weekly conference calls were helpful in
keeping the preparation laboratory operating in
an efficient and consistent manner by allowing
potential difficulties to be identified, discussed,
and  resolved.   The preparation laboratory
manager, EMSL-LV QA staff, ERL-C sampling
QA staff, and SCS state office representatives
participated in these calls. Issues discussed
during  the conference  calls included  field
supply needs, clarification of procedures,  e.g.,
sample labeling, and  shipment of samples.
Routine calls, however, were more beneficial
for dealing with specific  issues.

Preventive Maintenance

     All laboratory equipment was maintained
in  accordance with  guidelines in the PLSOP
(see Appendix A). The analytical balances and
                                           15

-------
pH meters were calibrated at the beginning of
every day of use.    The instruments were
handled and stored according to the guidelines
identified  in  the manufacturers' instruction
manuals.  The convection ovens, muffle fur-
naces and walk-in cooler were monitored  to
ensure proper functioning. An air compressor
was used to clean the surfaces of the sample
preparation  equipment,  e.g.,  rolling pin and
sieves,  after processing each  sample.    A
desiccant trap device was built into the airflow
conduit  to collect  any  vapor or  oily liquid
residues before reaching the nozzle. In addi-
tion  to  cleaning with  compressed air,  the
processing equipment was wiped with clean
tissues  if soil was  adhering to the surfaces.
The  sample  drying and sieving areas were
swept   and  damp-mopped  once  per week
during the exchange of samples  in the drying
area.  A  laboratory safety plan  was imple-
mented successfully during the survey.

Special Studies

     Although there were no deviations from
the  specified protocols for  sample  drying,
concerns  were  raised in the previous DDRP
surveys about soil homogenization techniques,
possible dust contamination of the samples,
and  the potential for leaching of chemicals
from the kraft paper and nylon mesh into the
samples.  Separate  studies  were undertaken
by the QA staff to investigate the effects of
these factors on bias   or variability  in  the
resulting data.  Copies of the internal  reports
summarizing these studies are found in Appen-
dices D, E, and  F, respectively.  Brief sum-
maries  of the special studies are provided in
the following subsections.

Soil Homogenization Study

     During the planning stages of the MASS,
it was decided that a study should be initiated
to assess the best methodology for  sample
homogenization  and subsampling (see Appen-
dix D).   The cone-and-quarter technique and
the  closed-bin  and open-bin riffle  splitting
techniques were compared for four soil sam-
ples of different  texture.  Particle size analysis,
organic matter  by loss  on ignition, and pH in
0.01M calcium chloride  were  used to  assess
differences  in replicate  subsamples from  the
different  techniques.   The  closed-bin  riffle
splitting technique using five complete  passes
through the device was determined to be the
optimal method.

Dust Study

     An external  dust  study was  initiated
before and  during the  construction  of the
sample drying  area  in the  warehouse (see
Appendix  E).   The study  was  intended  to
assess the amount and constituents of dust
collected in the general warehouse facility and
in the enclosed sample  drying  room.  The
study was conducted prior to the receipt of
samples for the survey.   It was concluded
that the possibility of significant dust con-
tamination in the  warehouse  was  low, par-
ticularly in the drying area.  As additional
safeguards against dust contamination, a pair
of enclosed ventilated rooms were constructed
for sample drying and sieving.

Paper and Mesh Study

      Another study assessed the possibility of
contaminants leaching into the samples from
the kraft paper and nylon mesh used  during
sample drying (see Appendix F).   It was con-
cluded that the mesh could not have served as
a contaminant  of  the samples,  but  that the
kraft  paper  could  have had the  potential  for
detectable contamination by sulfur oxides and
sodium when a moist or wet sample made
contact with the paper.

      To investigate this latter possibility, three
low-range audit samples that were identical to
the low-range laboratory audit (LAL)  samples
used  in the survey were  saturated with deio-
nized water and were air-dried in the sample
drying area using the same  procedure that
was used for the routine samples.  Each of
these  samples, coded  "PAL" in  the  MASS
analytical data base, were placed in a batch of
samples  sent  to  a  separate laboratory  for
analysis.  The data for the PAL samples were
compared to the LAL accuracy windows and to
the LAL in their respective batches (see Appen-
dix F).  In the first batch, the PAL was within
the accuracy windows for all parameters. The
PAL in the second batch satisfied the criteria
for all parameters except  the sodium extracted
in calcium chloride. This parameter was above
the upper boundary of the  window by 0.001
meq/100g. Within these two batches, the LAL
contained more values that did not satisfy the
accuracy criteria  than the  PAL sample.  The
                                            16

-------
third batch contained five parameters in which
the PAL samples were above the upper boun-
dary of the accuracy window.  Three parame-
ters of  particular concern  were sodium ex-
tracted with calcium chloride, sulfate extracted
with deionized water, and the sulfate adsorp-
tion isotherm.  However, the sodium value for
the LAL  was  also  above the maximum ac-
curacy window value, and  reanalysis of the
extractable sulfate parameter proved that the
original value was erroneous.

     It  was determined that no  significant
differences existed  between the LAL, which
was not placed on the  kraft paper, and the
PAL.  This  finding  suggests  that  the  kraft
paper  was   not  contributing  a  significant
amount  of sulfur or sodium contamination to
the MASS samples.

Data Quality Assessment

     A primary goal of the QA program  was
to provide a detailed description of measure-
ment quality in accordance with objectives that
were defined  prior  to beginning the survey.
The following  subsections  describe the dif-
ferent phases  of assessment.

Quality Attributes

     The  preparation laboratory data  were
assessed in  accordance with the  following
quality attributes:
  • Precision - The  level of agreement among
    multiple measurements of the  same soil
    characteristic.
  • Accuracy - The level of agreement  bet-
    ween  an observed value  and  the "true"
    value of a  soil characteristic.
  • Representativeness - The degree to which
    the soil data collected accurately represent
    the population of interest.
  • Completeness - The quantity of data that
    is successfully collected with respect  to
    the amount intended in the experimental
    design.
  • Comparability -  The similarity of data  from
    different sources included in a given set of
    data, and  the  similarity of  methodologies
    from related projects across the regions of
    interest.
Measurement Quality Objectives

     To address the overall DDRP objectives,
conclusions must  be based on scientifically
sound  interpretations of the data base.  Ac-
cordingly, the EPA requires all monitoring and
measurement programs to  have  established
objectives for data quality based on the an-
ticipated end uses of the data (Stanley and
Verner,  1985; USEPA, 1986).  The utility of the
data is defined by the ability to confirm, reject,
or discriminate among hypotheses formulated
in the various DDRP computer  models. The
primary purpose  of the QA program is  to
increase the likelihood that the resulting data
base meets or exceeds the overall data quality
objectives (DQOs).  The actual quality of data
can be quantified in relation to  the  DQOs,
thereby allowing the data user to evaluate the
hypotheses with a known level of  confidence.

     In practice, DQOs are  statements of the
levels of uncertainty that a data user is willing
to accept  in the  data.   For the MASS QA
program,   measurement  quality  objectives
(MQOs) were defined as surrogates for user-
defined  DQOs.  The MQOs  for the sampling,
preparation, and analysis phases  were speci-
fied in  the  MASS QA Plan (Papp et al., 1989).
The  initial  sample  preparation  MQOs  were
established in accordance  with the require-
ments of EPA data users and were based  on
the selection of appropriate methods to obtain
the necessary data. The MQOs were reviewed
by scientists familiar with soil preparation and
analytical methods, including soil chemists and
laboratory  personnel.  Modifications  to the
preparation laboratory MQOs and protocols
were implemented using information  gained
from the DDRP  Northeastern region and Sou-
thern Blue  Ridge  Province surveys and from
peer review comments, in accordance with the
limitations of a particular procedure or analyti-
cal instrument.

     Soil sampling in the MASS included the
physical removal  of soil samples  from  an
excavated pit as well as the characterization
of the  soil pedon and  sampling site.  The
MQOs developed for sample preparation and
analysis were not applicable to the field sam-
pling activities; hence, specific sampling objec-
tives were developed  to  ensure that  field
operations  were conducted  in  a consistent
manner  and that  an estimate  of variability
among sampling crews could be provided. The
                                           17

-------
primary goal of the  MASS sampling was to
describe and collect  soil samples from repre-
sentative pedons of  15 established sampling
classes (Lee et al., 1989).  The sampling clas-
ses were intended to represent the range of
soil types encountered in the Mid-Appalachian
region.   The field sampling effort  produced
qualitative soil data from pedon characteriza-
tion, e.g., soil color, and quantitative data from
soil analysis, e.g., soil pH.

     To  assess  the  sample   preparation
MQOs, a series of measurement quality sam-
ples were analyzed together with the routine
samples in  a manner that  was  statistically
relevant and in which conclusions concerning
the data quality could be made.  There are a
number of assumptions upon  which the run/
sample design was based:
  • Complete attainment  of  the analytical
    MQOs could be accomplished if each run
    of   samples  satisfied  the   acceptance
    criteria.
  • The components of measurement uncer-
    tainty at any given phase,  e.g., within-run
    precision, can be evaluated with data from
    a  minimum  of  20 measurement quality
    samples of a given type.
  • The primary components of measurement
    uncertainty to be  assessed at the prepara-
    tion  laboratory were  precision and  ac-
    curacy.

     Table  3-1 presents the  MQOs for  the
preparation  laboratory parameters of interest.
A brief  description of the  attributes as they
relate to the achievement of MQOs follows.

     Precision is estimated  in this report  by
calculating  the  standard  deviation  (SD)  or
relative standard deviation  (RSD) from  the
QE/QC  samples for the  seven  parameters.
Additional precision estimates using analytical
data from  the  preparation  duplicates  are
documented  in the QA report for the  MASS
analytical  laboratory data (Byers et al.,  in
press).

     The accuracy characteristics of bias and
intralaboratory trends are  assessed  only for
the PH_MP and OM_LOI parameters because
reference values were not available for any of
the other  parameters.  The  MOIST_P values
were assessed to determine if the MQOs of
2.5 percent moisture for mineral and 6 percent
for organic soils were satisfied. A consistent
air-dry moisture value was expected because
the field audit material was well homogenized
prior to subsampling.

     The primary characteristics of represen-
tativeness at the sample preparation phase
are addressed in the QA report on the analyti-
cal laboratory data because the assessment
involves the  use of data from the preparation
duplicates  that  were  not analyzed  at  the
preparation laboratory.

     Completeness is measured by the labor-
atory's  performance  of  the  analyses  and
processing tasks for all samples satisfying the
sample receipt criteria.

     Comparability of the DDRP preparation
activities is evaluated using the precision data
for rock fragments and bulk density from the
different surveys.  The evaluation of prepara-
tion laboratory comparability  is  also  docu-
mented in  the  QA report on the analytical
laboratory data (Byers et al., in press).

Uncertainty Estimation for Laboratory
Analyses

     The QA staff was interested in assessing
the  various  components  of   measurement
uncertainty with respect to the  overall uncer-
tainty confounded in the routine sample data.
The following subsections describe statistical
and assessment procedures that were used.

Statistical Model

     Depending on its limitations or assump-
tions, each measurement  procedure induced a
random error for  each physical or chemical
characteristic of a soil sample.  The sum of
these errors can be defined as data collection
error, which  is treated as a  random variable.
This variable can  be  placed in  an  additive
model  (Cochran,  1977)   and serves  as   an
estimate  of  uncertainty  in a data set, such
that:
              y  =   n  +  e
where:  y is an observed sample characteris-
        tic,
        /j is the "true" sample characteristic,
        and
        e is the  data collection error,  which
        is  assumed to  be  the sum  of  the
        errors generated by the independent
        measurement phases.
                                           18

-------
Table 3-1.  Preparation Laboratory Wlthln-Run Measurement Quality Objective*
Parameter
PH_MP

OMJ.OI

MOIST.P

RF_FG
RF_MG
BD_CLD
BD_KV
Sample*
type
Buffer (QCCS)
Int dup'
FD
FAL(C)
FAP (B)
FAP (Bw)
FAO (O)
MS (Bs)
Int dup'
FD
FAL(C)
FAP (B)
FAP (Bw)
FAO (O)
MS (Bs)
Int dup*
FD
FAL(C)
FAP(B)
FAP (Bw)
FAO (O)
MS (Bs)
FD
FD
Reps'
Dups'
Reporting
format
0.01 pH units
II
II
II
II
II
II
0.01 wt. %

0.01 wt. %
II
II
II
0.1 wt. %
0.1 wt. %
0.01 g/cm3
0.01 g/cm3
Expected
range
3.90-4.10
2.50-7.50
2.50-7.50
4.80-6.00
4.40-5.40
4.70-5.70
3.70-4.50
4.10-5.00
0.00-90.00
0.00-90.00
0.00-3.00
5.00-15.00
1.00-9.00
65.00-85.00
5.00-15.00
0.50-6.00
0.50-6.00
0.00-2.50
1.00-4.00
1.00-4.00
4.00-8.00
0.50-6.00
0.0-60.0
0.0-60.0
0.60-2.00
0.40-2.50
Acceptable
accuracy
3.95-4.05
as specified
as specified
as specified
as specified
as specified
—
control limits
control limits
control limits
control limits
control limits
_
control limits
control limits
control limits
control limits
control limits
...
...
—
_
Acceptable
precision
0.15 SD
0.20 SD
0.15 SD
0.15 SD
0.15 SD
15% RSD
20% RSD
15% RSD
15% RSD
15% RSD
15% RSD
20% RSD
15% RSD
15% RSD
15% RSD
20% RSD'
20% RSD'
20% RSD'
20% RSD'
° Int dup = internal duplicate, FD = field duplicate, FAL = field audit low, FAP = field audit pair,
 FAO = field audit organic, MS = manager's sample.
* The internal duplicate is a replicate subsample that was randomly selected from one soil sample per analytical run.
' The replicate bulk density samples are clod samples in which one to three clods per horizon were sampled.
d The duplicate bulk density samples are known volume samples in which one or two samples per horizon were sampled.
' Variability was anticipated because these duplicates or replicates are measurements of different samples within the
 same horizon; however, when variability >20% RSD, these samples were checked for possible measurement error.
      The identification of the error distribution
throughout the measurement phases requires
a  large  number of  replicate measurements
which, from a budgetary and logistical stand-
point, can be restrictive. However, a relatively
small number of observations can be used to
estimate the first two moments, i.e., the mean
and variance, of the data collection error.  The
moments are  sufficient  to  estimate  accuracy
and precision of the routine data in an additive
model.  For this report, the  standard deviation
(SD) in parameter reporting units or the rela-
tive standard  deviation  (RSD), i.e., coefficient
of variation, in percent,  were used to assess
the precision components.  Bias was consid-
ered to be the quantitative measure used to
assess the accuracy component. The sum of
these error terms represents the measurement
uncertainty  in  a QA sample data sets,  and
includes both  measurement and population
uncertainty in the routine sample data set. In
order to make effective use of the  MASS prep-
aration laboratory data, the evaluation of mea-
surement uncertainty in relation to the overall
data uncertainty is essential.

     The expected  squared error was used
as an estimate of uncertainty and was calcu-
lated as the sum of the error variance and the
square of the bias term described  below.  The
error terms can be defined  only  for a given
analyte concentration range because the error
variance changes  with analyte  concentration;
however, the entire concentration range can be
                                              19

-------
partitioned into intervals and a step function is
applied to those intervals in such a way that
each step value is the squared error, E^e2), for
a specific interval of the entire concentration
range. Details of the  step  function  are  con-
tained in  the MASS QA report on the analytical
laboratory data (Appendix J in Byers  et at., in
press).

      For most data interpretation uses,  it is
preferable to convert tha squared uncertainty
values into the standard reporting units of the
parameters of interest. This is performed by
calculating the square root of the  squared
error.   To  evaluate the data uncertainty as-
sociated  with the measurement system and in
the population of  routine  data, the VEj(^)
terms  from each  interval  were pooled  and
weighted by the proportion of routine sample
concentrations within the corresponding inter-
vals.  The resulting  uncertainty estimate,  6 or
delta, is defined as:
         5   -  I rVE,(f?)  +  1 Pj
where:  5  is :n« pooieo uncertainly for a!i
        intervals over the concentration range,
        P,  is the proportion of routine sam-
        ples  in the  ith  interval, and
        E^e2)  is the mean  squared  error for
        the ith  interval,  and  represents  the
        sum  of the error variance  and  the
        squars of the bias for the ith interval.

      For the  BD_CLD and BDJ
-------
were too numerous to allow statistically-based
sampling of all soil components. As an alter-
native,  soils  were grouped  into  sampling
classes belie ^d to have similar physical and
chemical properties in relation to soil respon-
ses to acidic deposition. Field sampling sites
were randomly selected  frum  the  areas of
occurrence •:•>' each sampling class  because
the object e ol tns sampling was to charac-
terize sampan  ciasses rather than  individual
'.vatershedjj or soil types.  Prior to the WAGS,
:n«, ~'ii;a , -,v  h:;tc;og>3rh;i*v  of nhysicai and
chemical  ;oi spercses  for indHdua!  sampling
•Ji . s&s ?stv! '•; ' ,' b ">.'*"'cn wr"3S -.'/as not know:';.
         >'  .•..<.  lysis, ui  tne  two  previous
ji ;nP  surveys  however,  has shown  that
oiasiiiert'tsiii i;j«v.$n.ainfy in  the routine  sam-
ples is cor>.K"o, ^to/'y parafnetar end  quantitatively  com-
i',.-.red vv  ih", e."uf,si'>e of uncertainty stemming
      :j''iounded fi&id sampling and  prepara-
      ror,  T';are ara as many as 22 primary
      ' types associated with each  of the 15
      rrfj classes,  forming a total of 148 dis-
tinct  {>ai ^filing  c!ass/S"iori7on  groups.  Taole
3-2 pffusems the number  and percentrge of
primary  horiion types and   the number of
                                                Table 3-2. Primary Horizon Type® for Sampling
                                                         Class/Horizon Groups
sampling
           clause:-";;  associated   with   each
     The uncertainty estimates are generally
expected to increcse with increasing sources
of confounded error.   For this report, two
types of deits values have been calculated: 83
and 84 values. The S( and 82 values could only
be calculated using data from the LAP/LAL and
PD  sarfipws v;hk;S', were not analyzed at the
("""epacati'ito laboratory.  ITia  83 values were
caiculetrd '-ssing data rrom the field duplicate
sampieK  •;; :he bulk density replicates, and t;
-------
Table 3-3.  Samples Placed In Mineral and Organic
         Soil Batches
Samples per batch
Sample description Mineral Organic
FD Field duplicate sample
FAL Field audit sample, low-range
FAP Field audit samples, pair
FAO FielJ audit samples, triplicate
PD Preparation duplicate samples
LAL Laboratory audit sample, low-range
LAP Laboratory audit samples, pair
LAO Laboratory audit samples, triplicate
OCA Quality control audit sample
RS Routine samples
1
1
2
0
2
1
2
0
1
30
1
0
0
3
2
0
0
3
1
30
audit samples were stored in amber Nalgene
bottles identical to those used for the routine
samples and were not opened or disturbed in
any manner  at  the preparation laboratory.
Within each batch, the preparation laboratory
manager randomized the samples and labeled
each with  a DDRP Label B (see Appendix A).
The only exception to  the random placement
was the  QCA sample  which  was  always
designated the 15th sample in a batch.

     Due  to a limited volume of some  audit
sample materials at EMSL-LV,  e.g., the Bw
horizon audit  sample,  the volume of  sample
provided for each audit sample was some-
times  less than that provided for the  associ-
ated routine samples in the batch. Therefore,
the identity of these audit samples may not
have been as  anonymous  to  a laboratory
analyst as originally intended. An attempt was
made to disguise low volume audit samples by
purposely  placing a few low volume routine
samples within these batches.

     At the preparation laboratory, there were
two distinct groups of measurement quality
samples used in the sample  analysis: quality
evaluation (QE) samples and quality control
(QC) samples.  The QE samples allowed an
independent assessment of data quality by the
QA staff,  while the QC samples enabled the
laboratory to  control measurement error.  In
order  to  assess  the  MQOs, both  types of
samples  were  analyzed randomly  with the
routine samples in a manner that was statisti-
cally relevant and in which conclusions con-
cerning the quality of data  could be made.
During analysis at the preparation laboratory,
an attempt was made to place each type of
QE/QC sample within  each  analytical  run.
However, due to the priority of the creation of
batches for the contract analytical laboratories
and  the  fact that not all  types of QE/QC
samples were available at the time of analysis,
some analytical runs did not contain the full
suite of QE/QC samples defined below.

Quality Evaluation Samples

     The preparation laboratory QE samples
are those samples which were submitted blind
to the laboratory staff, i.e., the laboratories did
not know the predetermined concentration of
the samples.  The QE  samples  provided an
independent  check on the  analytical process
and were used at the preparation laboratory to
evaluate whether the MQOs set forth in the QA
Plan  (Papp et al.,  1989) were satisfied for all
analytical runs of samples.  Every QE sample
had a specific purpose in the data assessment
scheme and each was  similar to the routine
samples  both  in matrix and in analyte  con-
centration. The QE natural soil audit samples
had been previously characterized by a suffi-
cient number of analytical laboratories to allow
the definition of  accuracy windows used in
data  assessment. The QE samples processed
and  analyzed  at  the preparation laboratory
included  field duplicate samples, field audit
samples, and a laboratory manager's sample,
as described below.

Field Duplicate Samples - One horizon  per
sampling crew for each third day of sampling
was  sampled in duplicate using  an alternate
trowel-full method as  specified  in  the  field
sampling protocols (Appendix A  in Kern and
Lee,  in press).  The first  sample of the pair
was  considered the routine sample and the
second sample  was  considered  the  field
duplicate. Both samples underwent the same
preparation steps  as all  associated routine
samples. Data from the field duplicate/routine
pair  allowed an estimate  to be made of the
combined uncertainty due to field sampling
and  sample preparation  and  was used to
assess precision within each analytical run.

Field Audit Samples - Three  types  of  field
audit samples were used  in  the evaluation
of measurement quality for mineral  and or-
ganic sample  runs of  PH_MP, OM_LOI, and
MOIST_P. At least one pair of  FAP samples
and  associated FAL sample were placed in
each run of mineral samples. Where possible,
                                           22

-------
a triplicate of FAO samples was placed in the
organic sample runs.  The FAP/FAL samples
were processed by the sampling crews when
sampling  every third mineral soil pedon, and
the FAO samples were processed whenever an
organic soil pedon was sampled.  The sam-
pling crews treated the samples as if they had
just been  obtained  from  the  excavated pit,
hence, the samples were sieved and bagged at
the time of sampling. The field audit samples
were then handled  as if they were  routine
samples and underwent all of the subsequent
soil preparation  and  analysis steps.  The
samples were used  to assess accuracy and
precision within each analytical run.

Manager's Sample- A manager's sample was
included  in each  analytical run of  PH_MP,
OM_LOI, and MOIST_P. The sample consisted
of natural audit material having a characterized
accuracy window for each parameter, although
the concentration range was unknown to the
preparation laboratory analysts.  The mana-
ger's sample was used to assess  accuracy
within  each run.   The samples  were imple-
mented into the  program in  October 1988,
approximately six weeks after sample prepara-
tion was underway, as an  additional check on
data quality.

Quality Control Samples

     The  preparation laboratory QC samples
are those  samples which  were known to the
laboratory staff and were  used to control the
quality of  analytical data produced.   The QC
samples analyzed at the  laboratory included
liquid QC check samples and internal duplicate
soil samples, as described below.

Quality Control Check Samples - Before begin-
ning a run of  pH analysis, a  quality control
check sample (QCCS) of pH 4.0 was analyzed
using a different batch of National Institute of
Standards and Technology (MIST) stock buffer
solution than that used for the meter calibra-
tion. Subsequently, after every 10 samples and
after the  final  sample of the run,  another
QCCS of pH 4.0 was analyzed. The value for
each QCCS was required to be 4.00 ± 0.05; if
not,  the electrode was  recalibrated according
to the guidelines in the PLSOP. This procedure
was followed consistently during the operation
of the laboratory.
Internal Duplicates - Within each analytical
run of samples for the PH_MP, OM_LOI, and
MOIST_P analyses, a composite internal dupli-
cate subsample of a routine bulk soil sample
was obtained and analyzed for every tenth
sample in the run.  The mean value of  the
internal duplicate  sample and its correspond-
ing routine sample was  required  to  show a
standard deviation or a percent relative stan-
dard deviation less than that specified in Table
3-1.  If these criteria  were not satisfied,  the
data were reviewed by the laboratory manager
and  QA manager, and a decision was then
made whether to re-analyze any of the sam-
ples.

Technical  Systems Audits

     Formal  technical systems audits of  the
preparation  laboratory were conducted  im-
mediately before the initial receipt of samples
and  twice during  the operation.  The audit
team consisted of the QA officer, QA manager,
and other  representatives of the EMSL-LV QA
staff and  representatives of the ERL-C sam-
pling QA staff.

     During  the  pre-operational  audit con-
ducted  in September 1988, the audit team
toured the laboratory facilities, including  the
sample drying room and the mobile laboratory
units.   During the first  operational audit in
October 1988, the audit team observed bulk
sample preparation  including the  sample
receipt and storage, sample drying and analy-
sis, disaggregation and sieving, and soil homo-
genization   and   subsampling  operations.
During the second operational audit  in March
1989,  the  team  evaluated  the bulk density
analysis and reviewed the concerns addressed
in the previous audits. Overall, the  systems
audits  did not identify any major flaws in  the
methods or operations used at the preparation
laboratory facility.   The  audit reports  are
presented  in Appendix G.

     A second type of  systems  audit was
instituted in which an EMSL-LV QA staff repre-
sentative visited  the  preparation  laboratory
facility approximately every two weeks. This
allowed minor difficulties to be resolved before
they had an adverse effect on the preparation
activities.
                                           23

-------
Data Verification

     A computer program was developed that
allowed entry and verification of preparation
laboratory data from the raw data forms. The
program also tracked soil samples through the
various sample processing and analysis  ac-
tivities,  and  assigned  verification  flags  to
discrepant data.  A remote modem assembly
allowed concurrent data assessment  by  the
QA staff.   A  summary of the preparation
laboratory verification procedures and flags is
presented in Appendix B.

     After the laboratory analysts entered raw
data on  the forms,  an  evaluation  procedure
was initiated which identified coding discrep-
ancies and data that did not satisfy the MQO
criteria previously outlined in  Table  3-1.  The
flags  identified in Appendix B were used to
qualify the discrepancies.  All  outliers were
confirmed or corrected before the data were
assigned to the verified preparation laboratory
data base.

     The summary  data  form  (DDRP Form
101) for each batch was reviewed by QA staff
to ensure that the sample codes  corresponded
to those entered on the field data and raw
data  forms.    The inclusion  of appropriate
QE/QC samples was checked and recorded for
each batch.  Analytical data were checked for
completeness and correctness.  The prepara-
tion  laboratory manager  was  instructed  to
acknowledge and correct any  errors in  the
data  while  reviewing the  raw data  forms,
control charts, or logbooks. The batch sample
shipping  form (DDRP Form 102) was compared
with the summary data  form  to  verify  the
analytical laboratory destination and the num-
ber of samples  shipped.  The data base mana-
ger  was notified  by the QA  staff  of  any
changes  made to the data base as a result of
these checks.

      Data verification and corrective action for
the preparation laboratory was  accomplished
primarily through  the application  of a  QA
reanalysis template (QART) for each parameter
of interest  (see  Appendix H).   The  QARTs
provided an assessment checklist of QE/QC
samples for each  analytical run  and were
designed to assess data according to prede-
fined  precision  limits and  accuracy windows.
The templates  were used to  make decisions
on whether to  request  reanalysis of a par-
ticular parameter.

     The QARTs were constructed as spread-
sheets,  where the row  entry fields specified
the evaluation criteria and the column entry
fields represented the severity of the measure-
ment uncertainty that had  occurred. The first
column  was used to identify major measure-
ment uncertainty  and  the second column
identified  minor uncertainty.   When  major
uncertainty occurred, the QA manager informed
the preparation laboratory manager to check
for errors in the data for the suspect run and
parameter before initiating reanalysis.  If the
laboratory confirmed the suspect values, then
reanalysis was requested for all  samples in
the run. The following types of measurement
uncertainty  were  considered  major  for  all
parameters  measured  at  the  preparation
laboratory  and  constituted the basis for re-
questing confirmation/reanalysis:
  • Precision/accuracy   -   The pair of FAP
    samples and the FAL sample (or the tripli-
    cate of FAO  samples) were  grouped to
    allow  three evaluations.   The FAPs and
    FAQs  were evaluated  for precision and
    accuracy, and the FAL was evaluated for
    accuracy.  Major uncertainty existed if any
    two of  these three evaluations failed to
    satisfy the MQOs.
  • QC check sample   -   Major uncertainty
    existed  if  a QCCS  exceeded the estab-
    lished control limits.
  • Cumulative minor uncertainty  - Three or
    more  occurrences  of  minor  uncertainty
    were considered to be equivalent to major
    uncertainty.

     The following types of measurement un-
certainty were considered  minor for all para-
meters  measured  at the preparation  labora-
tory:

  • Precision/accuracy  -  Minor uncertainty
    existed  if  one of  the three  precision/
    accuracy evaluations  did not satisfy the
    MQOs, or if the FD, internal duplicate, or
    manager's  sample  did not satisfy  the
    MQOs.

     For each QART, a table was  constructed
to summarize the QE/QC  samples analyzed
within  an analytical  run.   For each type of
sample, an  asterisk next to a precision value
or the letter "N" designates a failure of that
                                           24

-------
value to satisfy the specific MQO.  The final
two columns,  one for the  QA manager and
another for independent  observation by a
representative  of the Environmental Research
Center, University of Nevada at Las Vegas,
were used to  designate pass  or fail for  the
run.  Although some runs  did not contain all
possible types QE/QC samples, it  was con-
cluded that all runs passed the QART criteria.
Tables 3-4, 3-5, and 3-6 summarize the QART
data,  by run,  for the PH_MP, OMJ.OI,  and
MOIST_P parameters, respectively. ~
     Although the verification programs iden-
tified values exceeding the MQO  criteria for
precision and  accuracy, trends in the data
could not be discerned.  Control charts were
used to detect trends, such as the data from
the B and Bw horizon FAR samples for PH_MP.
It  was determined that  a  pH reference elec-
trode was replaced about the time that  the
shift was detected (approximately October 30,
1988). Control charts spanning the duration of
the MASS sample preparation are presented in
Appendix I.
                                           25

-------
Table 3-4. Summary of Achievements for Acceptance Criteria of Field-Moist pH (PH_MP)
Date of
run
9/17/88
9/20/88
9/21/88
9/23/88
9/24/88
9/28/88
9/29/88
10/1/88
10/4/88
10/6/88
10/8/88
10/13/88
10/15/88
10/19/88
10/22/88
10/27/88
10/31/88
11/3/88
11/7/88
11/11/88
11/15/88
11/19/88
11/23/88
12/3/88
12/10/88
12/15/88
12/20/88
12/28/88
1/3/89
1/10/89
FAP
[acc/prec]
(0.15 SD)
—
—
—
Y
0.08
Y
0.00
YY
.06 .05
—
YY
.04 .06
Y
0.04
Y
.01
YY
.04 .01
Y
.04
Y
.00
Y
.08
YY
.01 .01
Y
.01
Y
.04
YY
.01 .00
YY
.02 .01
YY
.01 .03
YY
.00 .02
YY
.01 .01
Y
.04
YY
.06 .09
Y
.04
YY
.01 ,01
Y
.01

Y
.01

FAL
[ace]
—
—
—
Y
Y
YY
—
YY
N*
Y
YY
Y
Y
Y
^Y
Y
Y
YY
YY
YY
YY
YY
Y
YY
Y
YY
Y

Y

QCCS
[ace]
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Int Dup
[prec]
(0.15 SD)
0.03
0.00
0.26*
0.07
0.04
0.06
0.06
.00 .01
.00
0.01
0.68*
.02 .01
.08
.04
.06 .01
.00
.02 .01
.01 .01
.01
.01
.02 .09
.01 .01
.05 .00
.00
.16* .01
.04 .01
.00 .01
.04 .01
.11 .09
.12 .00
.01 .02 .01
.04 .03
.01 .02
.03 .00
.05 .00 .06
.02 .01

.00 .06 .05
.01 .06
.06
.03
.01 .00
.03
.01 .00
.00
.01 .01
.02 .10
.00 .01 .01
.01 .01 .01
.01 .01
.12 .00
.00 .01
.01
FD
[prec]
(0.20 SD)
—
—
—
0.02
0.00
0.00
0.03
0.02
0.00
0.08
0.01
0.03
0.03
0.09
0.03
.29*
.13
.06
.04
.03
.03
.13
.07
.08
.01
.08
.04
.07
.16
.11
.04
.01
.03
.07
.06
.10

MS
[ace]
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Y
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
Y
Y
Y
Pai
LESC
Y
Y

Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
B
ERG
Y
Y

Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
                                             26

-------
Table 3-5. Summary of Achievements for Acceptance Criteria of Organic Matter (OM_LOI)
Date of
run
9/20/88
9/23/88
9/27/88
9/28/88
9/30/88
10/3/88
10/7/88
10/8/88
10/14/88
10/15/88
10/19/88
10/22/88
10/25/88
10/28/88
11/2/88
11/8/88
11/10/88
11/15/88
11/18/88
11/23/88
11/29/88
12/09/88
12/10/88
12/15/88
12/19/88
12/22/88
12/23/88
12/31/88
1/6/89
1/14/89
1/21/89
FAP
[acc/prec]
(15% RSD)
—
—
Y
1.0
Y
2.4
YY
3.7 0.0
YY
1.6 1.6
Y
3.8
Y
0.01
YY
0.8 1.5
—
Y
1.6
Y
2.2
YY
4.2 0.9
YN
39.6* 5.1
Y
1.2
YY
1.0 2.3
Y
1.9
YY
1.6 0.2
YY
0.7 0.9
YY 6.8
Y 0.9 1.5
Y
0.2
-
Y
5.8
Y
5.0
YY
2.1 1.6

Y
0.8

Y
3.0


FAL
[acc]
—
—

Y
YY
YY
Y
Y
YY
—
Y
Y
YY
YY
Y
YY
YY
YY
YY
YYY
Y
-
Y
N
YY

Y

Y


Int Dup
iprecl
5% RSD)
—
—









3.5 2.8
18.4* 1.1
0.1 2.9
2.4 1.1
0.2 1.6
7.5
0.3 1.4
4.0 1.9
0.4
2.0
2.4
2.0 6.5 8.1
7.0 1.9
11.6 2.4
10.7 1.3
1.0 1.8
0.5 7.4
21.7*
1.9 2.4
8.9
2.9
11.0
1.4
2.5 5.1
5.6
8.3 12.6
0.9
0.5
2.7 0.3
2.4
1.1 1.8
2.4 13.6
0.6 5.4
16.3* 6.9
2.1
4.1
2.1
FD
(20% RSD)
—
—
4.7
17.5
4.6
7.1
1.2
1.6
0.8
1.8
8.4
11.6
14.3
2.1
0.3 0.4
2.0
15.8
3.6
0.8
0.7
0.6
2.7 0.1
3.0
17.7
12.4
13.6
4.2
24.6* 7.7
2.4
3.7
1.1
2.6
5.9
3.8
11.5 5.7

1.0 0.3
2.4

10.9


MS
[acc]
—
—










y
Y
Y
Y
N
Y
y
y
Y
y
y
Y
y
N
y
y
Y
Y
Y
Pa
LESC


Y
Y
Y
Y
Y
y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
y
Y
y
y
y
Y
Y
y
18
ERC


Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y

27

-------
Table 3-6. Summary of Achievements for Acceptance Criteria of Air-Dry Moisture (MOIST_P)
Date of
run
9/20/88
9/22/88
9/24/88
9/27/88
9/28/88
10/1/88
10/7/88
10/8/88
10/12/88
10/18/88
10/21/88
10/22/88
10/25/88
10/29/88
11/3/88
11/7/88
11/11/88
11/15/88
11/19/88
11/23/88
12/3/88
12/10/88
12/15/88
12/20/88
12/28/88
1/3/88
1/10/89
1/18/89
FAP
[acc/prec]
(151 RSO)
—
—
—
Y
1.3
YY
1.5 1.1
YY
6.1 119*
Y
5.6

YYY 3 . 7
2.7 3.2
Y
1.5


YY
0.6 0.6
YY
4.6 2.2
Y
0.0
YY
3.6 2.1
YY
1.5 0.8
YY
1.4 0.8
YY
2.6 4.3
Y
0.5
Y
9.8
Y
4.5
Y
0.4
YY
6.6 5.4
Y
6.3

Y
3.4

PAL
[ace]
—
—
—
Y
YY
YY
Y

YYY
Y


YY
YY
Y
YY
YY
YY
YY
Y
Y
Y
Y
YY
Y

Y

Int Dup
[prec]
(151 RSD)
—
—
—
—








3.0 10.2
2.3 7.7
3.0 3.2
3.2 7.3
43.6*
3.3
3.3
1.0
3.6 4.9
3.4
7.1 2.1
0.5 0.6
4.4 9.7
1.7 1.1
4.9
3.1
20.3* 0.3
1.4 2.4
11.2 3.9
8.1
4.0
2.4
0.3 6.1
1.3
15.7* 3.1
2.7
0.6 13.9
5.7
2.9 9.3
2.8
3.0
7.2
FD
[prec]
(20% RSD)
—
—
—
9.0
2.3
6.0
2.9

2.1
8.5
19.1


14.8 3.8
0.8
23.6*
20.8*
12.1
17.4
14.4
7.8
3.5
23.9*
7.1
1.6
32.2*
4.5
10.1
3.2
2.2
10.9
6.5
18.7

24.0*

MS
[ace]
—
—
—











Y
—
Y
Y
—
Y
Y
Y
Y
Y
Y
Y
Y
Y
Pae
LESC



Y
Y
Y
Y
8
ERC



Y
Y
Y
Y
ONE SAMPLE
ANALYZED
Y
Y
Y
Y
TWO SAMPS
ANALYZED
THREE SAMPS
ANALYZED
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
                                         28

-------
                                      Section 4

                           Results and Discussion
     This section provides  the  results and
discussion for data quality assessment of the
preparation laboratory data.   Achievement of
the MQOs is expressed  in  relation to  the
actual  MQOs specified in the MASS QA Plan
(Papp et al., 1989).

Evaluation  of Quality Attributes

     The preparation laboratory  data  were
assessed in a systematic manner using statis-
tical results  from the measurement quality
samples.

Precision

     With the exception of  the bulk density
and rock fragment data, it  was  determined
that concentration  dependency  would  not
affect the presentation of pooled  imprecision
estimates. The precision information for the
field audit samples are presented individually
for each horizon type, thereby eliminating the
concentration dependency issue.   The  field
duplicate and internal duplicate observations
were  within  very narrow  ranges, therefore,
precision was not strongly affected by  con-
centration.

     Statistical  analysis of  precision   was
performed for the PH MP, OMJLOI, MOIST P,
RF_FG, RF_MG.  BD_CL.D, and BD_KV parame-
ters. Because bulk~density and rock fragment
data displayed some concentration dependen-
cy, these data were evaluated separately for
values  above and below the mean.  A root
mean  square error  statistic was  used  to
estimate the  pooled standard deviation  (SD)
or,  where  appropriate,  the   pooled  relative
standard deviation (RSD) across each labora-
tory data set.  The  percent  RSD  was deter-
mined by dividing the SD by  the mean of the
data set, and multiplying this quotient by 100.
Field-Moist pH (PH_MP)

     The  pH parameter was included in the
MASS for  two reasons:  (1) to compare field-
moist soil  pH values with air-dry soil pH values
that were  obtained during contract laboratory
analysis,  and  (2)  to  indicate the possible
presence  of  inorganic carbon.   Rather  than
perform an inorganic carbon test on every bulk
sample  as was done  in the  NE and SBRP
surveys, the test was required only on sam-
ples of pH 7.0 and higher.

     Summary statistics of  the  pooled  im-
precision  estimates for  the measurement
quality samples are provided in Table 4-1. The
imprecision estimates for analysis of field-
moist pH  are well within the specified MQOs
for the different measurement quality samples
that were  analyzed. The data did not exhibit
concentration  dependency  because  the  pH
values occupied a narrow range of concentra-
tion.

Table 4-1.  Imprecision Estimates for PH_MP
Sample
type
FD
FAP (Bw)
FAR (B)
Int Dup

n-pairs
34
20
19
92

Mean
4.84
5.14
4.93
4.93
SD MQO
— pH units —
0.08 0.20
0.03 0.15
0.03 0.15
0.05 0.15
Organic   Matter  by
(OMJLOI)
Loss   on  Ignition
     This parameter was measured to distin-
guish more  accurately organic soils  from
mineral soils. Organic and mineral soils were
placed in separate batches to expedite analy-
sis at the analytical  laboratories.  All samples
containing 20 percent or more organic matter,
even though identified in the field with mineral
                                           29

-------
soil designators, were placed in organic soil
batches.

     Summary statistics  of  the  pooled im-
precision  estimates  for  the  measurement
quality samples are provided in Table 4-2.  The
imprecision estimates for  analysis of organic
matter are well within  the  specified  MQOs.
The higher RSD value for the FAR Bw  can be
explained  by  one outlier pair.   This pair  had
individual values of 8.34 and 4.69, with the 8.34
value  being suspect.  Due to a lack of avail-
able sample in archive,  this sample was not
reanalyzed although the raw data sheets and
entry were scrutinized for transcription errors.
If these two  samples are removed from the
summary calculation, the RSD for the remain-
ing  19 pairs is 1.63 percent.
Table 4-2. Imprecision Estimates for OMJ.OI
Table 4-3.  Imprecision Estimates for MOIST_P
Sample
type
FD
FAR (Bw)
FAR (B)
Int Dup
RSD MQO
n-pairs
46
20
19
64
Mean
9.02
4.72
10.16
7.56
— percent
6.68
12.99
3.48
9.48
	
20
15
15
15
Air-Dry Moisture (MOIST_P)

     Summary statistics  of  the pooled im-
precision  estimates  for  the  measurement
quality samples are provided in Table 4-3.  The
average RSD for the FD/routine pairs is greater
than the MQO.  However, the only requirement
in  the  PLSOP was that the  laboratories dry
samples  to less than 2.5  percent for  mineral
and 6 percent for organic samples.  Once this
was accomplished the samples  were con-
sidered ready  for further processing.   For
example, a routine sample containing a mois-
ture content of 1.0 percent and its associated
field duplicate containing  a moisture content
of 2.0 percent  were both  acceptable even
though the pair had a calculated  RSD of 47
percent.

Rock Fragments (RF_FG and RF_MG)

     Summary statistics  of  the pooled im-
precision estimates for rock fragment content
are  provided in Table  4-4.  Since  the audit
Sample
 type
n-pairs
Mean
 RSD    MQO
— percent —
FD
FAR (Bw)
FAP(B)
Int Dup
46
20
19
64
1.85
1.84
3.37
1.09
21.57
3.73
4.36
11.71
20
15
15
15
Table 4-4. Imprecision Estimates for RF_FG, RF_MG
Parameter
RF FG"
> mean
< mean
RF MG°
> mean
< mean
n-pairs
46
24
22
46
24
24
Mean
7.86
12.50
2.85
15.74
27.25
5.66
RSD MQO
— percent —
19.04
13.76
42.69
16.61
13.33
22.87
20


20


                                               " Estimates based on field duplicate/routine pairs.
samples  had already  been sieved, the only
samples for which imprecision could be mea-
sured were the FD/ routine pairs.  The inten-
tion was to identify only those extreme outly-
ing values where  sampling or  preparation
errors might have occurred.

     The  RSD  values less than and greater
than the  mean were evaluated in  order to
determine whether or not  the statistical  rela-
tionship of higher RSDs at  lower  concentra-
tions, in this  case  at lower rock fragment
content, would hold true  for rock fragment
data.  The RSDs presented in the table appear
to confirm this relationship.  Because of the
simplicity of the method used for determining
the percentage of rock fragments  in the bulk
soil samples,  a  greater  amount of the im-
precision can be  attributed to spatial within-
horizon  variability or sampling  error and  a
lesser amount to  preparation laboratory error.

     Data for  the two  rock fragment  size
classes were quite variable, especially at low
concentrations  of rock  fragments.   For ex-
ample, one 20-mm rock fragment in one of the
two samples compared (when both samples
had low rock fragment contents) could  show
an inordinate degree of variability.  This situa-
tion  is  not unusual when  considering the
                                            30

-------
natural spatial heterogeneity of soil horizons.
Any pairs having larger than 20 percent RSD
were checked for transcription errors and the
fragments were  reweighed.

Bulk Density (BDjCLD and BDJCV)

     Summary  statistics of the pooled  im-
precision estimates for bulk density are provi-
ded in Table 4-5.  Imprecision was determined
using replicates of each  horizon  sampled.
Clods were usually sampled in triplicate while
known volume samples were usually sampled
in duplicate.  Due to the highly variable nature
of the bulk density samples, i.e.,  either struc-
turally intact  or reconstituted, the preparation
of an audit sample was not possible.  There-
fore, imprecision was estimated solely on the
replicate data.
Table 4-5. Imprecision

Parameter n-pairs
BD CLD" 705
< mean 347
> mean 358
BD KV* 145
< mean 75
> mean 70
Estimates

Mean
1.51
1.28
1.74
0.78
0.42
1.18
for BD_CLD, BD_KV
RSD MQO
— — osrcsnt -----
7.52 20
10.04
5.59
17.52 20
23.77
14.36
° Estimates based on clod replicates.
* Estimates based on known volume replicates.
     The  RSD  values  less than  and greater
than the  mean were  evaluated  in  order  to
determine whether or not the statistical rela-
tionship of higher RSD at lower concentrations,
in this case at lower bulk densities, would hold
true for bulk density data.  The percent RSDs
presented in the table  appear to  confirm this
relationship. It is also  evident that the known
volume sample mean values are considerably
lower than the clod mean values and that the
variability  is higher.  The lower means were
due to the collection of known volume samples
from horizons containing more organic matter,
unnatural voids, or rock fragments, all of which
generally  tend  to lower  bulk  density.  The
techniques used  in collecting known volume
samples are also considered to be  less pre-
cise than the technique used  to collect clods.
     The data suggest a consistency of bulk
density values within a given horizon.  Audit
reports indicated  that the sampling  crews
were able to select representative clods from
each horizon and that  the preparation labora-
tory   was   consistent  in  its measurement
techniques.  However,  the exact  percentage
of error in the data due to horizon variability,
sampling error, or laboratory error cannot be
determined   because:    (1) inherent  within-
horizon spatial variability  made it  improbable
that identical clods could  be sampled, or that
a  clod audit sample for  measurement of
potential sampling error could be provided, and
(2) audit samples that could theoretically allow
the estimation of preparation laboratory error
were not provided  to the laboratory.

     The following  types of  possible  field
sampling errors  could  affect the data for the
bulk density replicates:
  • Collection of replicates from transitional
    zones or adjacent horizons,
  • Mislabeling  of clods,
  • Inconsistent saran coating procedure, or
  • Spatial variability relating to the coherency
    of clods.

     Based on the experience level of the SCS
sampling crews  and the fact that  the various
field audit reports did not indicate major devia-
tions from the protocols, sampling error is not
presumed to have been  a  significant factor
affecting the RSD  values.  Some variability in
the use of  saran was noted, although  the
imprecision estimates would not be affected  if
the coating  procedure  was consistent for all
replicates within a horizon.

     Measurement uncertainty may be intro-
duced at the preparation laboratory because of
measurement or method errors, such as:
  • Transcription  errors, e.g., misrecorded
    weights or sample codes,
  • Inconsistent saran coating procedures,
  • Improper clod handling, e.g., compaction,
  • Incomplete  drying,
  • Loss of soil or rock fragments during
    sieving, or
  • Faulty weighings, e.g., balance  not calibra-
    ted or tared.
There was no evidence to suggest that any
such errors  occurred during the preparation
laboratory operations.
                                            31

-------
Accuracy

     The  audit  samples used to  assess
accuracy were natural soil samples. Although
their "true" chemical composition and physical
characteristics were unknown, accuracy  win-
dows were established using soil  analytical
data from the previous DDRP surveys and by
preliminary analysis of new parameters added
to the preparation procedures. Accuracy was
based upon the data for FAP, FAL, FAO, MS,
and QCCS samples.  Tables  3-3, 3-4,  and 3-5
in Section 3  summarized the acceptance or
failure of the samples  to satisfy the  sample
run acceptance criteria.  The audit samples
were also tracked on control charts to deter-
mine if certain trends could be detected  (see
Appendix I).

     Accuracy could not be assessed  for rock
fragment or  bulk density data.  As the audit
samples were initially  prepared and  sieved,
there were no rock fragments in the samples
except  those  that entered  the sample  via
contamination from sampling or preparation.
Also, there were no  audit samples used for
bulk density analyses.

Field-Moist pH (PH_MP)

     The accuracy windows for PH_MP were
derived  from  confidence intervals   placed
around a biweight  estimation (Kadafar, 1982)
of the mean.  In almost all cases, the within-
run inaccuracy for all audit samples satisfied
the MQO for the respective accuracy windows.
Only one FAL outlier and two MS outliers were
recorded.  A  trend towards higher  pH values
was noticed in both the B and Bw horizon FAP
samples on or about October 30,  1988  (see
Appendix I).   It was initially suspected that a
change  in electrodes or a drop in ambient air
temperature might be responsible for such a
shift, although neither  possibility  was  con-
firmed.  However, the trend was not noticed in
the manager's sample  nor the FAL  sample.
The increase  was not great enough to cause
the FAP samples to exceed the limits of their
respective accuracy windows for PH_MP, and
was well within the range of expected error for
multi-phase pH systems (Brezinski, 1983).

Organic Matter Loss-on-Ignition (OMJ.OI)

     The accuracy windows for OMJ.OI were
based  on a 15 percent RSD  confidence limit
placed about the mean values of ten audit
samples  at  each concentration.   Except for
one FAP,  one FAL, and one MS, the inaccuracy
estimates for the audit samples satisfied the
QART criteria for each analytical run.

Air-Dry Moisture (MOIST_P)

     The initial accuracy criteria for MOIST_P
were  that the  values  should  be below 2.5
weight percent for mineral soils and 6.0 weight
percent for organic soils.  All audit samples,
with the  exception of the  manager's sample
and the B horizon FAP sample, satisfied these
criteria.  Both types of samples were allowed
to dry for six days, i.e., twice the normal drying
period, with no significant change measured in
their mean moisture content of approximately
2.3  and   3.3  weight  percent,  respectively.
Control charts with 10 percent RSD  warning
limits and 15 percent RSD control limits around
the mean values were defined for the samples.
Using these different criteria, all audit samples
satisfied  the accuracy objectives for air-dry
moisture.

Representativeness

     Representativeness  is addressed in the
QA report on the analytical data (Byers et al.,
in press) by using data from  the preparation
duplicates to assess the subsampling proce-
dure performed by each laboratory.

Completeness

     The requested analyses and soil proces-
sing steps were performed on 100 percent of
the  bulk  samples that satisfied the  sample
receipt and batching criteria. Samples that did
not satisfy these criteria included some groups
of field  duplicates  and audit samples that
were not sampled according to the field sam-
pling  protocols  or that were extraneous in
terms of the proportion of duplicate samples
and  audit samples to the total number of
batches created.  Preparation  duplicates were
created from the field duplicate/routine sample
pair in each  batch sent to the  analytical labor-
atories, for  a  100 percent level of complete-
ness.  Hence, the MQO for completeness at
the preparation laboratory was satisfied.

     Of  the 844 routine horizons  sampled
in the field, 705  horizons were  sampled for
clods, resulting in 83.5 percent completeness.
                                           32

-------
Therefore, the  completeness  objectives  for
sampling were  satisfied.  Only three of  the
2,021 clod samples received at the preparation
laboratory could not be analyzed, resulting in
99.9 percent completeness.

Comparability

     Most of the preparation laboratory com-
parability issues are evaluated in the QA report
on the analytical laboratory data (Byers et al.,
in press).  However, it was possible to inves-
tigate and  compare  the imprecision  results
from data collected throughout the three DDRP
surveys  for those parameters that were com-
mon to  the  surveys.  Only  two parameters,
rock fragments and clod  bulk density, were
analyzed in all three surveys. Table 4-6 sum-
marizes  the data from an inter-survey com-
parison  cf  the  mean  percent RSD  pooled
across laboratories for the  two parameters.
Clearly,  the  data  are very  comparable and
probably reflect the  best precision that can
be achieved with the specified methods.  It
should be noted that the variability in the NE
survey rock fragment data is probably under-
estimated because of differences  in the field
duplicate sampling method (Byers et al., 1989).
Certain  sampling  crews  in the  NE  survey
created a "split" duplicate sample, while others
sampled the duplicate using  the  (intended)
alternate trowel-full method.
Table 4-6. Multiple Survey Comparison of Imprecision
         for Rock Fragments and Clod Bulk Density
Pooled RSD°
Parameter
Rock fragments*
Clod bulk density5
< mean
> mean
NE
13.21
8.32
8.97
6.38
SBRP
14.83
10.84
14.72
7.61
MASS
17.83
7.76
10.04
5.59
' Pooled relative standard deviation, in percent, across
 laboratories (NE=4, SBRP=2, MASS=1).
* Data from field duplicates, including both fine and
 medium gravel combined.
' Data from clod replicates.
Assessment of Data Uncertainty
Components

     The  following   subsections  describe
various data uncertainty components that are
estimated using data from the measurement
quality samples.  The effect of measurement
uncertainty on overall data uncertainty is also
discussed.

Additive Components of Uncertainty

     As described previously in Section  3, the
within-run imprecision estimates were pooled
over the concentration range using the propor-
tion, P(i), of routine data  in each interval to
derive the within-run imprecision component of
the delta values.  Also, it was assumed that
the laboratory analyzed the soil samples on a
run-by-run basis under conditions specific to
each run's analysis.  Hence, the between-run
imprecision was calculated for the  PH_MP and
OMJ.OI parameters using the data from the
FAR samples. Bias for these parameters was
calculated from the  FAL  samples  and  the
mean values  of the FAR sample  pairs.  The
within- and between-run variances were added
to the squared  bias term and were converted
into reporting units to produce delta values for
each  data set  and  parameter.   Table  4-7
shows the imprecision and bias estimates and
the contribution of the between-run  and bias
components in  relation to the within-run im-
precision estimates for the analytical  com-
ponent.

     The within- and between-run mean sums
of square and the squared bias component of
analytical  uncertainty  can  be experimentally
compared with  the three mean squares of a
randomized block design (Box et al.,  1978).  In
this scenario, both the  expected value of the
squared bias and expected value  of the bet-
ween-run mean square includes the expected
value of the within-run mean square  as calcu-
lated from the field audit samples. Ratios are
used  to  highlight those  parameters  which
indicate a  considerable contribution of  the
                                           33

-------
Table 4-7. Comparison of Measurement Uncertainty Components
Parameter
PH MP
OMJ.OI
Reporting
units
pH units
wt. %
Within-run SO
—..— units — — —
0.0285
0.5484
Between-run SD
	 linjtg 	 .««.
0.0842
0.2619
Bias
- units -
0.0119
0.5223
Ratio
B-R/W-R
3.0
0.5
Ratio
Bias/W-R
0.4
1.0
NOTE: Comparisons were not applicable for MOIST_P, RF_FG, RF_MG, BD_CLD, and BD_KV parameters.
between-run  and  bias  components to  the
within-run component, as shown in the table.
If the between-run mean square is two times
larger than the within-run mean square, then
the between-run imprecision can be inferred to
be  as large as the  within-run  imprecision.
Similarly, if the  between-run mean square is
more than three times larger than the within-
run mean square, then  the between-run im-
precision can be inferred to  be significantly
larger than the within-run imprecision.

      Hence,  if the ratio of the  between-run
component with the  within-run component in
reporting units exceeds Vs, i.e., 1.73, then the
between-run  imprecision  can be considered
significant in relation to  the within-run impre-
cision.  The  same criteria  can  be used to
examine the significance of the bias in relation
to  the  within-run imprecision.    Under  this
framework, there is no evidence of significant
between-run or bias uncertainty if the ratio is
less than 1.  If the ratio is between 1 and 1.73,
the  observed  uncertainty is  negligible  and
could be due simply to random fluctuation. If
the ratio is greater  than 1.73, then the  bet-
ween-run or bias component could be as large
as 50 percent of the within-run component.

      The  ratio of between-run to within-run
was greater than 1.73  only  for  the PH_MP
parameter. Possible reasons for this parame-
ter  to show  substantial between-run error
could relate to the unique aspects of its meth-
od of analysis, e.g., composite subsampling of
a nonhomogeneous field-moist bulk sample, or
very precise  within-run  results  which  could
magnify the effect of  between-run imprecision
and yield a high ratio. The  PH_MP within-run
imprecision was so small, i.e., less than one-
sixth of the MQO, that the between-run com-
ponent appears to be substantial.  In actuality,
both types of uncertainty are very low. There-
fore, although the contribution of the between-
run uncertainty to total analytical uncertainty
might be  substantial,  the magnitude of the
uncertainty is relatively small.  The bias was a
relatively small component of uncertainty, and
all ratios of bias to within-run imprecision were
below the critical value of 1.73.

Effect of Measurement System on
Data Uncertainty

     Table 4-8 provides the S3 and 84 values of
data uncertainty for the preparation laboratory
parameters.   As expected,  the 83 estimates
were much lower than the 84 estimates, which
are based on the field duplicate and sampling
class/horizon data sets, respectively (NOTE:  8,
and 82 values could  not be calculated for the
preparation  laboratory  parameters).   The  83
values contain the confounded overall  mea-
surement  uncertainty of  field sampling  and
sample  preparation and analysis.    The  84
values contain uncertainty due to the spatial
heterogeneity of  the routine sample population
confounded  with the  overall measurement
uncertainty  of  field  sampling and sample
preparation and  analysis.
Table 4-8. Delta Values and Ratios for Assessment
         of Data Uncertainty Components
Parameter
PH MP
OM LOI
RF FG
RF MG
BD CLD
BD KV
Reporting
units
pH units
wt. %
wt. %
wt. %
g/cm3
g/cm3
Mean"
4.72
11.93
7.36
13.47
1.42
0.78
Delta values*
83 84
0.1154
0.7820
1.5575P
2.0666P
0.1125P
0.2837P
0.5047
3.5038
5.8436
11.8267
0.1824
0.7431
Ratio
63
1
•)
1
1
1
1
S4
4
4
4
6
2
3
NOTE: Delta values were not applicable for MOIST_P.
• Mean value of routine samples.
* 83 calculated from field duplicates and 84 calculated
 from sampling class/horizon groups of routine samples.
p Partial estimate because of incomplete information.
                                            34

-------
     It has been proposed that a  measured
value can be considered as essentially error-
less for most uses if the uncertainty  in that
value is one-third or less of the permissible
tolerance  for its use (Taylor, 1987). The QA
staff was interested in determining the effect
of measurement uncertainty on overall data
uncertainty in the routine samples.  The overall
measurement uncertainty was based on the S3
values  calculated  from the  field  duplicate
samples.  The magnitude of this uncertainty in
relation to the associated S4 data uncertainty
values  generated from the  sampling  class/
horizon groups of routine samples provides the
data user with a basis  for  assessing  the
contribution of the measurement system to the
data uncertainty.   For the purposes of this
assessment, the data uncertainty confounded
in the sampling class/horizon groups has been
considered to be a surrogate for user-defined
error tolerance values. Where 83 is one-third or
less of S4, measurement uncertainty is a negli-
gible contribution to the overall data uncertain-
ty.

     Under this   framework,  measurement
uncertainty is negligible for  all  parameters
except BD_CLD.  The relatively high measure-
ment uncertainty can probably be attributed to
naturally high spatial variability within a given
horizon in a given  pedon due  to the irregular
presence of rock fragments, pores, roots, or
localized compaction. Nevertheless, the mea-
surement uncertainty attributed to the data is
only a  fraction  of the  allowable  MQO  for
BD  CLD.
                                            35

-------
                                     Section  5

                   Conclusions and Recommendations
     The conclusions and recommendations
discussed below have been summarized from
information supplied by preparation laboratory
personnel,  QA staff, and  other  survey par-
ticipants. The recommendations are presented
for consideration and possible implementation
in future surveys.

Laboratory Operations

     The MASS preparation laboratory opera-
ted in a  very efficient manner and prepared
analytical samples of equal or higher quality
than those prepared  in earlier DDRP surveys.
This quality can be attributed to the centralized
preparation facility and to the many QA proce-
dures that  were implemented for the MASS,
including the ventilated sample drying room,
improved cleaning procedures, increased QA
oversight, and the skill level of the laboratory
personnel.

Laboratory Design

     The laboratory design greatly increased
the efficiency of the preparation operations.
Sample  receipt, storage,  and supplies  were
within the facility for quick access. The facility
was designed to allow samples to flow direct-
ly from one operation to another.

Equipment and Supplies

     Originally, the saran:acetone clod dipping
solution was mixed at the preparation labora-
tory and shipped to the sampling crews.  The
integrity of the mixture upon delivery to the
crews was not suitable.  Much of the acetone
apparently volatilized during transport, thereby
increasing the viscosity of the saran:acetone
solution.  It is recommended that pre- weighed
packets be sent to field crews who can then
purchase acetone locally and create adequate
supplies of the  mixture as needed.   Safety
instructions for the preparation of the mixture
should be provided.

Sample Receipt

     The  sample receipt  procedure greatly
improved detection of samples mislabeled in
the field. // AS recommended that an accurate
list of the watershed and pedon identification
codes be provided, as was done in this survey,
to the preparation laboratory manager prior to
sampling.   Laboratory personnel  discovered
that pre-labeled field  sampling labels, i.e., the
majority of sampling  information were coded
on the labels before sample collection, general-
ly contained fewer errors.  These labels were
usually more legible, making it easier to iden-
tify codes.  // is recommended that  the sam-
pling crews perform as much pre-labeling as
possible before going to the sampling site.

     Preparation laboratory  personnel were
available during  the  hours  of the  delivery
service (Saturdays included)  to  receive sam-
ples from  the sampling crews and  to check
the validity of the sample labeling using the
sample receipt criteria.

Sample Drying

     Due to the  low relative humidity and the
use of mesh drying tables, sample drying time
was minimal at the Las Vegas facility, thereby
improving  the efficiency  of  the preparation
laboratory operation.  However, it is recom-
mended that more attention be given to the
disaggregation of peds, especially in clayey or
organic soils, before they dry to a hard consis-
tency.  These types  of peds  were extremely
difficult  to disaggregate without using exces-
sive force.
                                           36

-------
Field-Moist pH

     This was the first of the DDRP surveys
to include field-moist  pH  as a  preparation
procedure. The pH parameter was included for
two  primary reasons:  (1)  to compare  field-
moist pH values with  values  that  were ob-
tained  on air-dry samples during contract
laboratory analysis, and (2) to use as a pos-
sible indicator  of the  presence  of  inorganic
carbon in a soil sample. Rather than perform-
ing an inorganic carbon test on every bulk
sample  (a procedure that  was performed in
the previous surveys), the test would only be
run on samples  a  field-moist pH of 7.0 or
higher. The most crucial step in the procedure
appears  to be the calibration of the pH instru-
ment. Strict adherence to calibration protocol
is required.   // is recommended that several
extra electrodes be available prior to sample
analysis.

Organic Matter by  Loss on  Ignition

     This was the first of the DDRP surveys
to include the determination of organic matter
content as a preparation procedure.  Organic
matter was measured in order to more clearly
distinguish organic  soils from mineral soils.
The  two  types  of  soils  were organized in
separate batches for more accurate analysis
at the contract analytical laboratories.  Sam-
ples containing greater than 20 percent organic
matter by loss on ignition, irregardless of their
field  horizon  designation,  were placed  in or-
ganic soil batches.

     There were no major difficulties encoun-
tered with the  procedure  used to determine
organic   matter.   When  analyzing organic
samples, it is recommended that only enough
samples be placed in the muffle furnace during
each run to prevent excessive  amounts of
smoke that accrues as the organic matter is
ignited.

Disaggregation and Sieving

     Determining the type and consistency of
soil material  that must be disaggregated is
important before initiating the procedure. // is
recommended that saprolitic material or soil
containing highly weathered rock fragments
should be disaggregated with extra caution to
prevent the accidental disaggregation of frag-
ments.
     It is recommended that the dispensation
of the rock fragments following analysis be
decided before soil processing begins.  This
action  will avoid lengthy or costly storage of
the fragments at the preparation laboratory.

Bulk Density

     The bulk density procedure was modified
slightly from the procedure  used in the two
previous surveys.  All clods were oven-dried
before  coating the  clods  with  the  saran:
acetone mixture at  the laboratory.   Drying
helped to  stabilize the initial weight and the
weight after clod dipping that, in past surveys,
fluctuated due to evaporation of moisture from
the clod during the clod dipping process.  A
2000-mL beaker is recommended for determin-
ing the clod weight in water.  Three to six dips
in a 1:5 by weight saran:acetone mixture is
sufficient  to seal a clod before submersion in
water.  It is recommended that the modified
method for submerging floating clods,  des-
cribed in Section 3, be added to the protocols.
It is recommended that the density of rock
fragments within each clod should not always
be assumed to be 2.65g/cm3, as is done many
bulk density studies.  In areas with a variety
of lithologies or weathering characteristics, an
appropriately lower  or  higher density value
could be used where necessary. An average
particle density  value of 2.47 was used for
clods  collected from  the  Mid-Appalachian
region.

Sample  Batching

     Care should  be  taken  to ensure  the
anonymity of measurement  quality  samples
within  a  batch.   These samples should  be
interspersed in the batch  and be identical in
weight and  appearance to routine samples.
The laboratory manager inspected and labeled
all samples in each batch.  // AS recommended
that all labels in a  batch be coded by the
same person, and typed if possible.

Sample Archiving

     Due  to the inadequate quantity of soil
material available  for  some  audit  sample
types,  the entire volumes of  the field audit
samples were used in the preparation of the
analytical samples. // is recommended that all
audit samples sent to the sampling crews be
of sufficient volume to ensure that a portion of
                                           37

-------
the sample can be archived upon receipt at the
preparation laboratory. This would allow the
QA staff to have a ready supply of additional
soil in the event that the analytical laboratory
depletes its supply.

Quality  Assurance Program


Training

     The preparation laboratory manager held
a training  session every two weeks to review
protocols as  well as help the laboratory per-
sonnel in their understanding of soil genesis
and morphology.  The sessions increased the
staff's appreciation for the work performed in
the field and  was a motivational strategy for
inviting participation in soil identification.

Measurement  Quality Objectives

     Because the  sampling crews  did  not
collect field audit and field duplicate samples
until the third pedon,  the initial runs of the
preparation laboratory parameters  did  not
contain all the samples necessary to evaluate
the measurement quality objectives using the
quality assurance reanalysis templates.  This
was allowed only because of time constraints
to prepare samples for contract analysis. // is
recommended that a full suite of measurement
quality samples be maintained in each run for
consistency in data quality assessment.  Field
audit and field duplicate  samples should be
processed with  the first pedon sampled and
every third pedon thereafter.  It is recommend-
ed that the number of internal duplicates be
decreased from one per ten routine samples to
one per run, as  was instituted at the DDRP
analytical laboratories.

      A two-tiered precision  assessment sys-
tem was used for the DDRP contract labora-
tory analysis parameters  in  which concentra-
tion dependent precision was evident.  Depen-
ding upon the data users'  requirements for
precision, two-tiered precision  objectives may
be suitable for the rock fragment  and organic
matter  parameters.  // is recommended that
the precision criteria for  air-dry moisture be
dropped.  Acceptance criteria should be based
solely on an allowable moisture content.
     Precision of the bulk density and rock
fragment data for the Northeastern region and
Southern Blue  Ridge Province  surveys  are
similar in magnitude to the Mid-Appalachian
survey data.

Technical Systems Audits

     The technical systems audit was a very
important phase of  the  quality assurance
program because it provided documentation as
well as  objective insights  to  the operation.
The  pre-operation  and  two systems audits
conducted during preparation operations were
helpful in clarifying the protocols and correct-
ing minor discrepancies.  // is recommended
that the quality assurance staff audits, which
were performed  approximately every two
weeks, be conducted on a less frequent basis.

Data Verification

     The Mid-Appalachian survey was the first
of the DDRP surveys which  utilized a  com-
puterized system for entry of all raw data.
The system greatly improved outlier detection
as  well as helping to determine progress  in
the field and in the  preparation laboratory.
Data entry was the responsibility of the labor-
atory manager and one  other  laboratory ana-
lyst. // is recommended that a person know-
ledgeable of the entry program,  other than a
preparation laboratory analyst, be responsible
for the data entry and verification activities.

     The immediate entry  of  sample receipt
data and a subsequent  verification of sample
codes  helped to identify miscoded  samples
within hours  of  sample  receipt.  The verifica-
tion program tracked the progress of all sam-
ples through the preparation and analytical
stages.  Upon entry of data for any particular
procedure or analysis,  the verification pro-
grams were run to identify any suspect data.
Any samples identified by this procedure could
be reanalyzed in the next run of samples. The
verification program improved sample flow and
provided a verified data base in a much shor-
ter timeframe  than  previous  surveys.   The
program also greatly increased the efficiency
of  outlier detection  and correction over  the
past surveys.
                                            38

-------
                                      References
Blake, G. R.   1965.   In   Black, C.  A.  (ed.)
      Methods of Soil Analysis, Part 1.  Ameri-
      can Society  of  Agronomy,  Madison,
      Wisconsin.

Brezinski, D.  P.  1983.   Kinetic, Static, and
      Stirring Errors of Liquid Junction Refer-
      ence Electrodes.  Analyst, 108, 425-442.

Byers, G. E., R. D. Van Remortel, J. E. Teberg,
      M.  J. Miah, C. J.  Palmer, M. L Papp, W.
      H. Cole,  A. D.  Tansey, D. L. Cassell, and
      P.  W.  Shaffer.   1989.    Direct/Delayed
      Response Project:  Quality Assurance
      Report for Physical and Chemical Ana-
      lyses  of Soils  from the Northeastern
      United States.  EPA/600/4-89/037.  U.S.
      Environmental  Protection  Agency, Las
      Vegas, Nevada.  216 pp.

Byers, G. E., R. D. Van  Remortel, M. J.  Miah, J.
      E. Teberg, M. L. Papp, B. A. Schumacher,
      B. L. Conkling, D. L. Cassell, and  P. W.
      Shaffer.   In press.  Direct/Delayed Re-
      sponse  Project:    Quality  Assurance
      Report for Physical and Chemical Ana-
      lyses of Soils from the  Mid-Appalachian
      Region of the United States.  EPA/600/x-
      xx/xxx.    U.S.  Environmental Protection
      Agency, Las Vegas, Nevada.  337 pp.

Church, M. R., K. W. Thornton,  P. W. Shaffer,
      D. L. Stevens, B.  P. Rochelle, G. R. Hol-
      dren,  M.  G. Johnson,  J. J. Lee,  R. S.
      Turner, D. L. Cassell, D.  A. Lammers, W.
      G. Campbell, C. I. Liff, C. C. Brandt, L H.
      Liegel,  G. D. Bishop,  D. C. Mortenson, S.
      M.  Pierson and D. D. Schmoyer.  1989.
      Future Effects of Long- Term Sulfur Depo-
      sition on Surface  Water Chemistry in the
      Northeast and  Southern Blue  Ridge
      Province: Results of the Direct/Delayed
      Response Project. EPA/600/3-89/061, U.S.
      Environmental Protection Agency, Wash-
      ington, D.C., 887 pp.
Cochran, W. G.  1977.  Sampling Techniques.
      3rd Edition.  J. Wiley & Sons, New York.
      428 pp.

Kadaf ar, K.  1982.  A Biweight Approach to the
      One Sample Problem.  Jour. Amer. Stat.
      Assoc. 77(378):416-424.

Kern, J. S., and  J. J. Lee.   In  press.  Field
      Operations and Quality Assurance Report
      for Soil Sampling in the Mid-Appalachian
      Region of the United States.  EPA/600/x-
      xx/xxx.   U.S. Environmental Protection
      Agency, Corvallis, Oregon.

Lee, J. J., D. A. Lammers, M. G. Johnson, M. R.
      Church, D. L Stevens,  D. S. Coffey, R. S.
      Turner, L. J. Blume, L. H. Liegel, and G. R.
      Holdren.  1989.   Watershed Surveys to
      Support an Assessment of the Regional
      Effect of Acidic Deposition on Surface
      Water Chemistry.  Environ. Mgt. 13:95-
      -108.

MacDonald, C. C.  1977.  Methods of Soil and
      Tissue Analysis in the Analytical Labora-
      tory.  Mar/times Forest Research Centre
      Information Report, M-X-78. Fredericton,
      New Brunswick, Canada.

Papp, M. L, R.  D. Van Remortel, C. J. Palmer,
      G. E.  Byers,  B.  A. Schumacher,  R.  L.
      Slagle, J.  E.  Teberg,  and M.  J. Miah.
      1989.   Direct/Delayed Response Project:
      Quality Assurance Plan for Preparation
      and Analysis of Soils  from the Mid-Ap-
     palachian Region of the United States.
      EPA/600/4-89/031.   U.S. Environmental
      Protection  Agency, Las Vegas, Nevada.
     225 pp.

Soil Conservation Service.  1984.  Soil Survey
     Laboratory Methods and Procedures for
      Collecting  Soil  Samples.  Soil Survey
     Investigations Report No.  1.  U.S. Gov-
     ernment Printing Office, Washington, D.C.
                                            39

-------
                                     References (continued)
Stanley, T. W.,  and S. S. Verner.  1985.   The
     U.S. Environmental Protection Agency's
     Quality Assurance Program,  in: Quality
     Assurance for  Environmental  Measure-
     ments.  ASTM STP 867. American Soci-
     ety for Testing and Materials, Philadel-
     phia, Pennsylvania,  pp. 12-19.

Steel, R. G. D.,  and J. H. Torrie.   1960.   Prin-
     ciples and  Procedures  of Statistics.
     McGraw-Hill Book Company, New York.
     481 pp.

Taylor, J. K. 1987.  Quality Assurance of Che-
     mical Measurements.  Lewis Publishers,
     Chelsea, Michigan.  328 pp.
U.S. Environmental Protection Agency.  1986.
     Development of Data Quality Objectives:
     Description of Stages 1 and II. Quality
     Assurance  Management  Staff.    U.S.
     Environmental Protection Agency, Wash-
     ington, D.C.  14 pp.

U.S. Environmental Protection Agency.  1989.
     An  Overview   of  the  Direct/Delayed
     Response Project. In: AERP Status, April
     1989.  EPA/600/M-89/005.  U.S. Environ-
     mental Protection Agency,  Washington,
     D.C.

Van Remortel, R. D., G. E. Byers, J. E. Teberg,
     M. J. Miah, C.  J. Palmer, M. L Papp, M.
     H. Bartling, A.  D. Tansey,  D. L Cassell,
     and P. W. Shaffer. 1988. Direct/Delayed
     Response  Project:  Quality Assurance
     Report for Physical and Chemical Ana-
     lyses of Soils  from the Southern Blue
     Ridge  Province of the United States.
     EPA/600/8-88/100.   U.S. Environmental
     Protection Agency, Las Vegas, Nevada.
     251 pp.
                                           40

-------
                            Appendix A

    Preparation Laboratory Standard Operating Procedures
               for the Mid-Appalachian Soil Survey
     The following protocols were used by preparation laboratory personnel during the Direct/
Delayed Response Project Mid-Appalachian Soil Survey.
                                  41

-------
         DIRECT/DELAYED RESPONSE PROJECT:
  PREPARATION LABORATORY STANDARD OPERATING
PROCEDURES FOR THE MID-APPALACHIAN SOIL SURVEY
                            by
          M. H. Bartling, M. L Papp, and R. D. Van Remortel
            Lockheed Engineering & Sciences Company
                   Las Vegas, Nevada 89119
                   Contract Number 68-03-3249
                       Project Officer

                         L. J. Blume
              Exposure Assessment Research Division
            Environmental Monitoring Systems Laboratory
               U.S. Environmental Protection Agency
                  Las Vegas, Nevada 89193-3478
        ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
             OFFICE OF RESEARCH AND DEVELOPMENT
            U.S. ENVIRONMENTAL PROTECTION AGENCY
                 LAS VEGAS, NEVADA 89193-3478
                            42

-------
                                        NOTICE
       This document is a preliminary draft.   It has not been formally released by the U. S.
Environmental Protection Agency.  It is being circulated for comments on its technical merit and
policy implications.

       Mention  of trade  names  or  commercial products does not constitute endorsement or
recommendation for use.
                                          43

-------
                                        ABSTRACT
       The Direct/Delayed Response Project soil surveys include the mapping, characterization,
sampling, preparation, and analysis of  soils as part of a national program to assess watershed
response to acidic deposition within certain regions of the United States. Soil samples from the
Mid-Appalachian region are being collected from predetermined sampling sites and are shipped to
a preparation laboratory for soil processing and preliminary analysis.

       The standard operating procedures contained in this manual will  serve as the protocols for
operations taking place at the preparation laboratory. The manual is submitted as a final draft on
September 25, 1988.
                                            44

-------
                                        CONTENTS
Notice  	  43
Abstract	  44

1.0  Introduction	  47
       1.1  Overview	  47
       1.2  General Laboratory Responsibilities	  49
       1.3  General Laboratory Support	  50
                     1.3.1 Facilities	  50
                     1.3.2 Equipment and Consumable Supplies 	51
                     1.3.3 Technical Assistance 	  52
                     1.3.4 Personnel	  53

2.0  Preparation Laboratory Logistics  	  55
       2.1  Sampling Equipment and Consumable Supplies  	  55
       2.2  Sample Identification  	  56
                     2.2.1 Types of Field Samples  	  56
                     2.2.2 Quality Evaluation Samples	  57
                     2.2.3 Quality Control Samples	  60
                     2.2.4 Use of the Sample Code  	61
       2.3  Sample Receipt	61
       2.4  Sample Storage   	  63
       2.5  Sample Tracking	  64
       2.6  Sample Batching	  65
                     2.6.1 Batch Configurations 	  65
                     2.6.2 Batching Procedure  	  66

3.0  Sample Processing and Analysis  	  69
       3.1  Sample Drying	  69
                     3.1.1 General Considerations	  69
                     3.1.2 Equipment	  70
                     3.1.3 Procedure	  70
       3.2  Field-moist pH Determination	71
                     3.2.1 General Considerations	71
                     3.2.2 Equipment	  72
                     3.2.3 Calibration and Standardization  	  73
                     3.2.4 Quality Control	  74
                     3.2.5 Procedure   	  75
       3.3  Organic Matter Determination by Loss on Ignition  	  76
                     3.3.1 General Considerations	  76
                     3.3.2 Equipment	  77
                     3.3.3 Quality Control	  77
                     3.3.4 Procedure   	  78
                     3.3.5 Calculations	  79
       3.4  Air-Dry Moisture Determination  	  80
                     3.4.1 General Considerations	  80
                     3.4.2 Equipment	  80


                                            45

-------
                                 CONTENTS  (continued)

                                                                                     Page

                     3.4.3 Quality Control	  80
                     3.4.4 Procedure  	81
                     3.4.5 Calculations	  82
       3.5  Disaggregation and Sieving	  82
                     3.5.1 General Considerations	  82
                     3.5.2 Equipment	  83
                     3.5.3 Procedure  	  84
       3.6  Rock Fragment Determination	  85
                     3.6.1 General Considerations	  85
                     3.6.2 Equipment	  85
                     3.6.3 Procedure  	  85
                     3.6.4 Calculations	  86
       3.7  Homogenization and Subsampling	  86
                     3.7.1 General Considerations	  86
                     3.7.2 Equipment	  87
                     3.7.3 Procedure  	  87
       3.8  Bulk Density Determination	  88
                     3.8.1 Clod Method   	  89
                     3.8.2 Known Volume Methods	  95

References	  98
                                            46

-------
                                      SECTION 1.0
                                     INTRODUCTION

       The Preparation Laboratory Standard Operating Procedures (PLSOP) serve as a guide for
personnel involved in the preparation of soil samples for the Direct/Delayed Response Project
(DDRP) Mid-Appalachian Soil Survey (MASS). All personnel must be trained by a laboratory
manager knowledgeable of the protocols detailed in this manual.
1.1  OVERVIEW

       The DDRP is an integral part of the Aquatic Effects Research Program (AERP) of the
U.S. Environmental Protection Agency (EPA).  The program is administered under the federally-
mandated National Acid Precipitation Assessment Program (NAPAP) to address concerns
relating to potential acidification of surface waters by atmospheric deposition within the United
States.

       The overall purpose of DDRP is to characterize geographic regions of the United States
by predicting the long-term response of watersheds and surface waters to acidic deposition.
The DDRP has been designed under the concept of regionalized integrative surveys.  According
to this concept, research programs initially are approached from a large region of  study leading
to the selection of regionally characteristic systems.  These systems can be assessed through
detailed, process-oriented research which will aid in the understanding of the underlying
mechanisms responsible for  observed  effects. The projected responses of watershed systems
typical of the regional population can then be extrapolated with confidence to a regional or
national scale.
                                           47

-------
       The EPA is assessing the role that atmospheric deposition of sulfur plays in controlling
long- term acidification of surface waters (EPA, 1985).  Recent trend analyses have indicated
that the rate of sulfur deposition is slowly declining in the northeastern United States and is
increasing in the southeastern United States.  If a "direct" response exists between sulfur
deposition and surface water alkalinity, then the extent of current effects on surface water
probably would not change much at current levels of deposition, and conditions would improve
as the levels of deposition decline.  If surface water chemistry changes  in a "delayed" manner,
e.g., due to chemical changes in the watershed such as increased cation activity due to mineral
weathering,  then future changes in surface water chemistry (even with level or declining rates of
deposition) become difficult  to predict. This range of potential effects has clear and significant
implications to public policy  decisions on possible additional sulfur emissions control (EPA,
1985b).

       Specific goals of  DDRP are to (1) define physical, chemical, and mineralogical
characteristics of the soils and to define other watershed characteristics across these regions,
(2)  assess the variability of  these characteristics, (3) determine which of these characteristics
are most strongly related to surface-water chemistry, (4) estimate the relative importance of  key
watershed processes in controlling surface-water chemistry across the regions of concern, and
(5)  classify the watersheds  with regard to their response to sulfur deposition and extrapolate
the results from the sample of watersheds to the regions of concern.

       The  MASS includes areas within the states of Pennsylvania, Virginia, and West Virginia.
The streams in this region were sampled as part of the National Surface Water Survey, which is
a NAPAP program designed and implemented by the EPA to conduct a chemical survey of lakes
and streams located in regions of the United States believed to  be susceptible to the effects of
acidic deposition.  A sampling design was applied to allow for unbiased  characterization of
regional populations, and resulted in the selection of 37 watersheds.
                                             48

-------
       Soil types that were identified in the Mid-Appalachian sampling region are combined into
groups, or sampling classes, which are either known to have or are believed to have similar
chemical and physical characteristics.  Each of the sampling classes is then sampled across a
number of different watersheds in which the classes occur.  Using this approach, a given soil
sample does not just represent the specific watershed from which it came.  Rather, each
sample contributes to a set of samples that collectively represent a specific sampling class for
all DORP watersheds within the sampling region.

       The preparation laboratory is an integral component of the MASS.  Personnel at the
laboratory are responsible for tracking soil  samples throughout processing, performing some
preliminary analyses, formulating and shipping batches of samples to the  analytical laboratories,
and documenting all activities.
1.2  GENERAL LABORATORY RESPONSIBILITIES

       The preparation laboratory is designed to be the link between the sampling crews and
the analytical laboratories. The primary functions of the preparation laboratory are to prepare
homogeneous, anonymous subsamples from processed bulk samples and to transfer batches
of those subsamples to the analytical laboratories.  For these tasks to be successfully
accomplished, the preparation laboratory must uniformly track, process, and store all samples.

       The preparation laboratory manager assumes the responsibility for maintaining the
integrity of all samples upon their arrival at the laboratory facility.  Samples are packaged and
shipped by the sampling crews utilizing an overnight express service.
                                           49

-------
       The preparation laboratory staff also performs certain soil analyses.  These include the
determinations of percent air-dry moisture, percent rock fragments in the 2- to 20-mm fraction,
percent organic matter content by loss-on-ignition, field-moist pH in water, and bulk density.

       After the samples have been processed and analyzed, they are  assembled in batches
and shipped to designated analytical laboratories for further analyses.  An important function of
batch formation is to place certain quality evaluation (QE) and quality control (QC) samples in
the batches without revealing their identity to the analytical laboratories.

       A separate function of the preparation laboratory is to distribute equipment and supplies
to the  sampling crews.  Through adequate tracking of supplies, the preparation laboratory can
avoid inventory shortages by notifying the EMSL-LV QA manager of any equipment needs.

       The preparation laboratory is required to keep thorough documentation of  sample
tracking and processing. This information must always be available to the EPA project officer
or designee throughout the project.  Documentation is submitted to the QA manager upon
conclusion of the survey.
1.3 GENERAL LABORATORY SUPPORT

1.3.1  Facilities

       The MASS soil preparation and preliminary analyses are performed in a secure, climate-
controlled warehouse located in Las Vegas, Nevada.  The warehouse contains a number of
mobile laboratory units that are used for specific preparation activities.  Fume hoods and
exhaust fans must be operable for certain stages of sample processing and for the preparation
of the clod dipping solution. A source of compressed air must be available  for routine cleaning
                                            50

-------
of equipment. Deionized water is required in the preparation laboratory for various analyses

and cleaning.  Cold storage facilities must be readily accessible with ample space to store at

least 600 bulk samples at a time.



1.3.2 Soil Processing Equipment and Consumable Supplies



       The following hardware and supplies are required for the prescribed soil processing and

analysis  procedures:
Log books
Raw data forms
Munsell color book
Pens, indelible black ink
Kraft paper, 36" wide rolls
Plastic gloves (unpowdered)
Dust masks
One-gallon paint cans
Aluminum weighing tins
Nalgene bottles: 2-L, 500-mL, 125-mL
Carboys, 13-gallon capacity
Crucibles, tolerance to  450°C
Evaporating dishes, tolerance to 450°C
Beakers, 1000-mL
Tongs
Ring stand
Thermometers, 0 to 100°C range
Sample drying tables, PVC / nylon mesh
Rolling pins, wooden
Rubber stoppers, No. 10 size
Brass sieves, square-holed: 2-mm and 4.75-mm
Riffle splitter, Jones-type, closed-bin, with 1.25-cm baffles
Top-loading balance, accurate to 0.01g, capacity to SOOg
Top-loading balance, accurate to 0.1g, capacity to 5Kg
pH meter, with proper electrodes
Desiccators
Desiccant
Hot plate
Convection oven
Muffle furnace
Shipping boxes
Packing material
Strapping  tape
Saran® powder
Acetone
Plastic containers: 25mL, 50mL
                                           51

-------
1.3.3 Technical Assistance

       Seven raw data forms have been designed by EMSL-LV for sample tracking and for
entering data relating to:  1) sample receipt, 2) bulk sample processing,  3) field-moist pH,  4)
air-dry  moisture,  5) loss-on-ignition, 6) clod bulk density, and 7) known  volume bulk density.
All raw data are recorded on these forms in indelible black ink and the completed forms are to
be organized in binders. An example of each form is provided in Appendix A

       A personal computer is available at the laboratory for entry and verification  of all
preparation data recorded on  the raw data forms. This provides an efficient way to track the
status  of the samples from the time of field sampling through shipment to the analytical
laboratory.  Through data comparisons, sample labeling discrepancies can be quickly identified
and reported to the Soil Sampling Task Leader at ERL-C and the QA Manager at EMSL-LV.
Final data values are calculated through a SAS-AF® computer program, allowing data to be
reviewed in real time.  Precision and accuracy of the QE/QC samples are  checked as the
samples are processed. This permits identification of any discrepancies and allows the QA
manager and laboratory manager to efficiently track progress of the laboratory  staff.

       The computer entry and verification system is capable of producing all of the final data
forms required by EPA. Selected laboratory personnel are trained to use  the computer system.
Appendix B describes the preparation laboratory data entry and verification system.

       Additional support is provided through communications with other DDRP personnel.  An
EMSL-LV QA representative periodically visits the preparation laboratory to aid in training
personnel and answering questions that arise. The laboratory manager, however, should never
hesitate to call the QA representative or other QA personnel if a problem  should arise.  The soil
sampling task leader at ERL-C and the QA manager are to be notified whenever there is a
question concerning the field sampling operation.
                                            52

-------
       All information pertaining to telephone conversations and written correspondence are to

be thoroughly documented in a log book.



1.3.4  Personnel



       The manager of the preparation laboratory is required to be knowledgeable in laboratory

methods and procedures, and have demonstrated ability to track large numbers of samples and

supervise laboratory personnel. Adequate staffing should be provided to ensure a fast and

efficient turnaround of samples from the field to the analytical laboratories.  All personnel must

be thoroughly trained in the protocols and procedures by the laboratory manager before the

processing of the samples begins.



       Ultimately, the laboratory manager is responsible for assigning duties according to the

specific project needs.  The following division of responsibilities is  tentative and may be

adjusted.



Laboratory Analyst A:

   o   Enters data  into computer entry and verification system
   o   Runs verification programs
   o   Receives  and logs samples at cold storage
   o   Organizes cold storage space
   o   Tracks all samples during  processing
   o   Assists other analysts after other duties are complete

Laboratory Analysts B, C, D, and E:

   o   Spreads samples to air dry
   o   Performs field-moist pH determinations
   o   Performs loss-on-ignition determinations
   o   Performs air-dry moisture determinations
   o   Disaggregates and sieves  bulk samples
   o   Performs rock fragment determinations
   o   Homogenizes bulk samples with riffle splitter
   o   Creates appropriate analytical and  archive subsamples
   o   Performs bulk density determinations
   o   Completes necessary  information on bulk sample processing sheet
   o   Assists other analysts after other duties are complete
                                           53

-------
Laboratory Analysts F and G [Part-time]:

   o   Disaggregates and sieves bulk samples
   o   Performs rock fragment determinations
   o   Tracks all DORP equipment (both field and laboratory)
   o   Coordinates equipment procurement, preparation, and distribution
   o   Assists other analysts wherever needed

Laboratory Manager:

   o   Coordinates  laboratory operations and time management
   o   Communicates with EMSL-LV QA manager and QA representative
   o   Communicates with ERL-C soil sampling task leader
   o   Oversees sample receipt and storage
   o   Oversees all computer data entry and evaluation procedures
   o   Oversees sample preparation and analysis activities
   o   Organizes analytical samples into batches
   o   Tracks all samples during processing
                                           54

-------
                                     SECTION 2.0

                         PREPARATION LABORATORY LOGISTICS
2.1  SAMPLING EQUIPMENT AND CONSUMABLE SUPPLIES



       A designated area within the laboratory area is used to store field sampling equipment.

The laboratory manager is provided with a master inventory list that is updated each time

equipment is distributed to the field.



       An initial package consisting of all sampling equipment and a portion of the necessary

consumable supplies is prepared and distributed to each sampling crew at the beginning of the

survey.  Additional supplies that  are anticipated  to be necessary for the completion of sampling

are periodically sent to the sampling crew leaders at the appropriate Soil Conservation Service

(SCS) field offices.  At a minimum, each initial field equipment  package includes the following:



Equipment/Consumables                                          Quantity

Hand pump                                                         1
35-mm camera w/built-in flash                                        1
Clod boxes                                                         30
Brass sieve, square-holed, 19-mm                                     1
Stapler, large                                                        1
Stapler, small                                                       1
Gel-pacs                                                           80
Coolers                                                            10
Hole punch                                                         1
Straight-edged trowel                                                1
Volume-fill 250-mL cylinder                                           1
Equipment for bead method                                          1
Field pH kit                                                         1
Random number table                                                2
SCS-SOI-232 field data form, dated 3/87                               50
Orange flagging                                                     10 rolls
Yellow  marker flags                                                  80
Indelible marker - thick                                               10
Indelible marker - thin                                                10
Golf tees (horizon delineation)                                        20

                                                                          (continued)


                                          55

-------
Equipment/Consumables                                          Quantity

Photogray cards - 8" x 10"                                           20
Horizon depth tape                                                  1
One-gallon metal can w/lid                                           2
Hairnets                                                           1 gross
Plastic inner sampling bags                                          350
Canvas soil bags                                                   350
DDRP Label A                                                      200
Staples                                                            2 boxes
Small plastic clod bags                                             600
Twist ties                                                         300
Clod tags                                                         300
Clod box labels                                                    500
Log books                                                        5
Saran® powder                                                    10 packets
Acetone                                                           5 gallons


      Supply  sheets are provided for the sampling crews.  Crews are requested to log all

consumables taken from or returned to the SCS offices, which allows better estimates to be

made of any additional consumable items needed. Log entries should be written in black ink

and identify the crew, the equipment taken, and the date.
       Equipment needs can be made known to the laboratory manager during weekly field

sampling conference calls.  Any items nearing depletion are promptly shipped to the appropriate

SCS office or another convenient delivery point. Each crew leader is responsible for the return

of sampling hardware to EMSL-LV after MASS sampling is completed.
2.2 SAMPLE IDENTIFICATION



2.2.1 Types of Field Samples



       Four types of soil samples are sent to the preparation laboratory during the field

sampling phase of the MASS.  These are: 1) bulk samples (routine and duplicate), 2) field audit

samples,  3) clod bulk density samples, and  4) known volume bulk density samples.
                                          56

-------
       Bulk samples are those soil samples taken from a designated portion of a specific
horizon.  A bulk sample contains approximately one gallon of soil and generally weighs about
5.5 Kg.  The bulk samples are deposited in large plastic bags and placed within protective outer
canvas bags.  Each bulk sample should arrive at the laboratory with a completed DDRP Label A
affixed to the inner plastic bag.

       Field audit samples are QE samples (described below)  sent by EMSL-LV QA staff to the
sampling crews for inclusion in the sampling of every third pedon by each crew. The samples
are sieved, packaged, and shipped in the same manner as their associated routine samples.
Each field audit sample should arrive at the laboratory with a completed DDRP Label A affixed
to the inner plastic bag.   Both mineral and organic horizons are sampled; for the MASS, organic
soils are defined as having 20 percent or more organic matter as determined by loss on ignition.

       Clod samples are fist-sized, structurally intact bulk density samples taken from
designated horizons.  Each clod is wrapped in a hairnet and dipped briefly in a 1:5
saranracetone solution to help maintain the clod structure and reduce moisture loss from the
clod during transport and storage.  Each clod is covered by a small plastic bag and placed in a
special clod box with dividers.

       Known volume bulk density samples  are taken from horizons where clods are
unobtainable.  There are two types of known volume samples being used in the  MASS:  volume
replacement (VR) and volume filling (VF) samples.  These bulk density samples are packaged in
small pre-labeled plastic bags in much the same manner as the bulk soil samples.

2.2.2 Quality Evaluation Samples

      The QE samples are those samples which are known to the preparation laboratory staff
but are blind to the analytical laboratory, i.e., the laboratory does not know the predetermined
                                           57

-------
concentration of the QE samples and cannot distinguish the QE samples from routine samples.
The QE samples provide an independent check on the analytical process and can be used to
evaluate whether the data quality objectives (DQOs) set forth in the DDRP-MASS QA Plan (Papp
et al., 1988) have been satisfied for any given batch or for all batches, i.e., overall measurement
uncertainty.  Every QE sample has  a specific purpose in the data assessment scheme.  The
samples are similar to the  routine samples both in matrix and in analyte concentration.

       The two types of QE samples are variability samples and accuracy samples. The QE
variability samples  are replicates of selected routine samples that are used to measure
precision of the analyses.  The QE accuracy samples  are audit samples that have been
previously characterized by a number of analytical laboratories, sufficient to generate accuracy
windows for data assessment.

2.2.2.1 Field Duplicate (FD) Sample --

       The FD is collected at  random  by sampling crews using alternate trowelsful of soil to
produce a pair of samples (one routine, one duplicate) from one specific horizon within a
specific pedon.  Individual  pairs are used to assess system-wide within-batch precision. These
estimates are pooled to provide the within-batch component of overall system measurement
uncertainty.

2.2.2.2 Field  Audit  (FAP/FAO)  Samples -

       The FAP and FAO samples are median-range audit samples sent in pairs or in triplicate
to the sampling crews, who treat the  samples as if they had just been obtained from the
excavated pit. The FAP samples are used when sampling every third mineral soil pedon and the
FAO samples are used whenever an organic soil pedon is sampled.  The field audit samples are
sieved and bagged at the time of sampling.  These samples are then handled as if they were
                                           58

-------
routine samples and undergo all of the subsequent soil preparation and analysis steps. The
FAP and FAO samples are used to identify system-wide bias within a batch.  The analytical
results are pooled from all batches to provide an estimate of all sources of measurement
uncertainty other than within-batch uncertainty.

2.2.2.3 Low-Range Field Audit (FAL) Sample -

       The FAL sample is a low-range audit sample sent to the sampling crews and is handled
in the same  manner as the FAP/FAO samples, but is included only in mineral soil batches. The
FAL is used  to identifv any system-wide contamination sources and as an additional check on
system bias. Pooled data from all FAL samples provide an estimate of the system detection
limit.

2.2.2.4 Preparation Duplicate (PD) Samples -

       The POs are split  samples of a field duplicate and its associated routine sample that are
created by the preparation laboratory staff.  The split samples are labeled PDF and PDR,
respectively.  The PD samples are used to identify the within-batch preparation component of
the overall measurement uncertainty and as independent comparative checks on the within-
batch precision of the analytical laboratories.

2.2.2.5 Laboratory Audit (LAP/LAO) Samples --

       The LAP and LAO samples are median-range audit samples identical to the FAP and
FAO samples except that they are sent in pairs or triplicate directly to  the preparation laboratory
for inclusion in the analytical batches.  The samples are used to assess analytical laboratory
bias within a batch and are used in combination with the FAP and FAO samples to identify
sources of bias,  e.g., contamination or analysis. Within each batch, the FAP/FAO and LAP/LAO
                                           59

-------
samples represent the same audit horizons (B, Bw, or 0 horizon). Data from the samples are
used to calculate interlaboratory differences and to assess between-laboratory and between-
batch components of measurement uncertainty.

2.2.2.6 Low-Range Laboratory Audit (LAL) Sample ~

       The LAL is a low-range audit sample identical to the FAL sample except that it is sent
directly to the preparation laboratory for inclusion in mineral soil analytical batches.  When
applied in conjunction with the FAL sample, the LAL is used to calculate instrument detection
limits (IDLs) as well as to identify the magnitude and, possibly, the source of sample
contamination.  The IDL is estimated as three times the standard deviation of the LAL samples
for a given analytical procedure.  Calculation of LAL data begins with the  second batch and the
results are used to independently check the laboratory detection limits to  ensure  that the DQOs
for detectability are being met.

2.2.3  Quality Control Samples

       The QC samples are non-blind samples established by contract to assist the
laboratories in meeting intralaboratory DQOs.  The quality control audit (QCA) sample  is the only
such QC sample batched at the preparation laboratory.

       The QCA is an audit sample provided to the analytical laboratories as the 15th sample in
every batch.  The laboratories are provided with the corresponding range  of reference values
obtained by previous DDRP multi-laboratory analysis.  The QCA sample is primarily used to
control bias and to reduce bet ween-laboratory and between-batch components of measurement
uncertainty. The QCA sample from each batch is tracked on a control chart to ensure that the
data are within acceptable  limits set forth in the analytical laboratory contract.
                                            60

-------
2.2.4  Use of the Sample Code

       Each soil sample must be identified by a sample code, which is a 13-character
designation uniquely identifying that sample.  The sample code also identifies those multi-bag
samples which are to be combined during soil processing.  The first three characters of the
sample code represent the type of sample, e.g.:
       R11  =  routine sample, single bag (R12/R22 or R13/R23/R33 for multiple bags, if
              necessary)
       FDD  =  field duplicate sample (FD1/FD2 for multiple bags, if necessary)
       FAL  =  field audit sample, low-range (mineral soils)
       FAP  =  field audit samples, pair (mineral soils)
       FAO  = field audit samples, triplicate (organic soils)
The next two letters designate the state:  PA for Pennsylvania, WV for West Virginia, or VA for
Virginia. Following this is a three-digit county designation.  A subsequent dash (or zero)
separates the county code from the pedon/horizon designation, which is the final four-digit
code.
NOTE: For the field audit samples, the last two digits of the sample code reflects the audit
sample type. For the  FAP samples, the B horizon samples contain the closing digits 01 and 02,
and the Bw horizon samples contain the closing digits 03 and 04. The FAO samples contain the
closing digits 01, 02, and 03.  The FAL samples always contain the closing digits 01.

2.3 SAMPLE RECEIPT

       The sampling crews should make every effort to inform the preparation laboratory when
samples are being shipped from  the field. Bulk soil samples, bulk density samples, and field
                                           61

-------
audit samples are to be placed in box-enclosed styrofoam coolers with two frozen gel-pacs.
The corresponding SCS-SOI-232 field data forms and sample shipment forms are  sealed in a
plastic bag and taped to the underside of the cooler cover. The outer box of the cooler should
be firmly enclosed with strapping tape. The samples are then shipped directly to  the
preparation laboratory via overnight carrier.

       Because of their fragile nature, the soil clods (one type of bulk density sample) are
securely packed with vermiculite in coolers separate from the bulk soil samples.  Clods are to
be stored in a well ventilated area while awaiting shipment.

       When samples are shipped on a Friday, the sampling crew must notify the preparation
laboratory manager to ensure that personnel are available on Saturday to receive  the samples.

       All preparation laboratory personnel must become familiar with the DDRP Label A (see
Figure  1), as it is a primary tool used in sample receipt and tracking.  There are several entries
on the sample label, including the date sampled, horizon designation, depth range of the horizon
sampled, crew ID, site ID, sample code, and set ID. The crew ID consists of four alpha-
numeric characters representing the state and the crew number assigned to each sampling
crew within the state.  The site ID is a code assigned by ERL-C  which designates a specific
watershed that is sampled.  The set ID is a five-digit code relating to each day a sampling crew
samples a pedon.  A range of set IDs are assigned to each crew.  The laboratory personnel
must also become familiar with the  DDRP Label B (see  Figure 2) used to label and track
analytical samples within the batches.

       The field data form is completed by the sampling crew at the sampling site. The form
lists all field data for a particular pedon, including descriptions of landform, vegetation, soil
climate, and a detailed pedon characterization. The field data form also contains codes, e.g.,
watershed ID, assigned by the soil sampling task leader.
                                            62

-------
              DDRP Label A
Date Sampled:
              D D M M M Y Y
Crew ID:
Site ID:
Sample Code:
Horizon:
Set ID:
                                Depth:	— _ _ _ ex
                Figure 1. DDRP Label A
      DDRP Label B

 Batch ID: 	

 Sample No: _ _
Figure 2. DDRP Label B
       The samples received at the laboratory are checked for coding accuracy against the field
data form and sample shipment form. A sample receipt raw data form (see Appendix A) must
be filled out completely for all samples received and logged-in at cold storage.  This information
is then entered and verified in the computer entry and verification system.  Samples should not
undergo processing until they are verified under this procedure.
2.4 SAMPLE STORAGE

       All samples are to be placed in cold storage as soon as possible after receipt. Known
volume bulk density samples, once checked through the sample receipt procedure, can be
placed in the drying area and allowed to air dry. Bulk soil samples and field audit samples
remain in cold storage until there is sufficient space in the drying area for the samples (grouped
according to set ID) to be air dried. After air drying, the samples are returned to cold storage
and must remain there when not being processed.
       The cold storage room must be maintained at a temperature of 4°C.  It is necessary for
the temperature to be monitored either by a continuous sensor or by daily thermometer
                                           63

-------
readings.  Any substantial deviations in temperature (± 2°C or more) should be recorded in a
log book; affected samples should be identified according to set ID and sample code.

       Samples should be organized in cold storage so that a particular sample within a given
set may be easily located. Samples arrive at the laboratory in sets which must be maintained
during storage, processing, and shipping.  Because organic and mineral analytical samples are
to be batched and shipped separately, it is advisable to keep organic samples separated from
mineral samples in the cooler.
2.5 SAMPLE TRACKING

       An adequate tracking system must be initiated to document the transfer of all samples
from the field to the preparation laboratory and, finally, to the analytical laboratory.  The
following protocols have been developed to ensure proper sample tracking.

       Before sampling begins, the ERL-C soil sampling task leader provides the  EMSL-LV QA
manager and preparation laboratory manager with a list of watershed identification numbers
(site IDs) and pedon numbers for each pedon to be sampled. This information is entered into
the computer entry and verification system  at the preparation laboratory before samples are
received.

       As the samples arrive at the  preparation laboratory, the sample receipt raw data form is
filled out and entered into the computer.  The information is checked against the  site IDs and
pedon numbers that have been entered into the program. If  a particular watershed/pedon
combination has been previously used, e.g., the preparation laboratory has previously received
samples with that watershed/pedon number or a particular watershed/pedon number cannot be
found on the ERL-C list, the soil sampling task leader and the QA manager are to be notified
                                           64

-------
immediately.  After the site IDs, pedon numbers, and other data on the sample receipt raw data

form are verified, soil processing and analysis can be initiated.
2.6 SAMPLE BATCHING



       After the sample preparation and preliminary analyses are completed for a number of

sets of samples, batches are constructed and readied for shipment to the analytical

laboratories. To preserve the integrity of each sample during these steps, careful guidelines

must be followed.



2.6.1 Batch Configurations



       A batch of samples assembled for chemical and physical analysis consists of a

maximum of 40 samples, including approximately 30 routine samples and 10 different QE/QC

samples which are used to assess or control various  types of measurement uncertainty.

Typical mineral and organic batches consist of the following sample configurations:
               SAMPLES PLACED IN MINERAL AND ORGANIC SOIL BATCHES

                                                             Samples per batch
Sample type         Sample description                      Mineral        Organic

FD                  Field duplicate  sample                     1              1
FAL                 Field audit sample, low-range               1              0
FAP                 Field audit samples, pair                   2              0
FAO                 Field audit samples, triplicate               0              3
PD                  Preparation duplicate samples              2              2
LAL                 Laboratory audit sample, low-range          1              0
LAP                 Laboratory audit samples, pair              2              0
LAO                 Laboratory audit samples, triplicate          0              3
QCA                Quality control  audit sample                1              1
RS                  Routine samples                          30             30
                                                            40 total        40 total
                                          65

-------
       Several sets are required to make up a batch.  A set should never be split across two
batches. If it is not possible to get 30 routine samples without splitting a set, as many
complete sets as possible are batched.  Within each batch, one set should contain  a field
duplicate and associated field audit samples.

NOTE: Organic samples are to be separated into unique batches that contain no mineral
samples, and vice versa. This is the only case where samples are separated from their
respective set.  All other batching and shipping procedures remain the same. An organic field
duplicate and associated field audit samples must also be included in the organic batches.

2.6.2  Batching Procedure

       Batch numbers and a preparation laboratory ID are assigned by the  QA manager for use
by the laboratory manager in batching samples.  All  batches must contain the types and
quantities of QE/QC samples described in Section 2.6.1.

       Within each batch, preparation duplicates are split from the field duplicate and its
associated routine sample. The sample code for a  preparation duplicate begins with the two
characters "PD" which identifies the sample as a  preparation duplicate.  The letter "F" in the
third space is used to identify the split from the field duplicate (PDF) and the letter "R"  is used
to identify the split from the routine sample (PDR).  The sample code continues with the final
ten alpha- numeric characters from the original sample, e.g., the split from routine sample
R11PA013-0405 would be designated as PDRPA013-0405.
                                           66

-------
       The sample code for a laboratory audit sample is comparable to that of its
corresponding field audit sample, with a letter "L" substituting for the letter "F" as the first
character of the code. The sample code for a QCA sample should read "QCAS", followed by a
dash and the 5-digit batch ID.  The LAP,  LAO, LAL, and QCA samples are provided to the
preparation laboratory directly from the EMSL-LV QA staff.  The samples are stored in amber
Nalgene bottles identical to those used for the routine samples.  The preparation laboratory
manager is responsible for labeling each of these  samples with a DORP Label B and  randomly
placing it within the proper batch.  The only exception to the random placement is the QCA
sample, which is always designated the 15th sample in every batch.

NOTE:  The audit samples are not to be opened or disturbed at the preparation laboratory for
any reason.  If the laboratory audit samples are packaged differently from the routine samples
upon their delivery to the preparation laboratory, e.g., with a different type bottle or cap, or with
unusual markings on the bottle, call the QA manager immediately for further instructions. The
QE samples must be blind to the analytical laboratory.

       With the exception of the QCA sample as the 15th sample, randomize all samples in the
batch and assign sample numbers from 1 to however many samples are in the batch, e.g.,  40
samples in a batch would be numbered 1 through  40.  Replace the  DORP Label A for each
analytical sample bottle with the appropriate DDRP Label B. Staple the DDRP Label A on the
corresponding bulk sample processing raw data form and initial the label so that the  writing
overlaps both the label and the form.

       Enter the analytical sample numbers into the computer batching program.  The program
then generates the DDRP Form 101 (see Appendix  A), including all column headings and final
data for the processed soil samples.
                                          67

-------
       Place the 500-mL and 125-mL analytical sample bottles in separate shipping boxes and
prepare a Form 102 (see Appendix A) for each box. Upon approval of the QA manager, ship
each box of samples to the specified analytical laboratory using an overnight express service.
The QA manager should  be provided with a clean copy of the Form 101 to serve as
documentation of the shipment.  Arrangements can then be made to  move the  2-L archive
sample bottles into long-term cold storage.
                                           68

-------
                                     SECTION 3.0
                          SAMPLE PROCESSING AND ANALYSIS

       Specific areas of the preparation laboratory are designated for sample processing and
analysis. During all of the steps of sample handling, maintaining the integrity each sample is of
the greatest importance.  Sample integrity is ensured by the use of detailed sample labels,  by
documenting the status of each sample during the processing and analyses, and by avoiding
physical or chemical contamination during  each processing step.

       The sample preparation protocols are described in detail throughout this section of the
manual.
3.1  SAMPLE DRYING

3.1.1 General Considerations

       Sample tables constructed of PVC and heavy nylon mesh are used to air dry the
samples.  Use of the mesh enhances air circulation and increases the rate of sample drying.

       Chemicals used in analyses are to be kept away from the drying area. Food, drinks,
and smoking are prohibited in the drying area.  A separate pair of gloves should be worn when
handling each sample.  The drying area should be damp-mopped at least once a week to
control dust.
                                          69

-------
3.1.2 Equipment

       Drying tables, PVC / nylon mesh
       Kraft paper, 36" wide rolls
       Plastic gloves (unpowdered)
       Bulk sample processing raw data forms

3.1.3 Procedure

       Label a bulk sample processing raw data form (see Appendix A) for each sample to be
air dried. Place two fresh sheets of kraft paper, approximately 1m x 1m in area, on the mesh
partition of the drying table. With gloved hands, slowly spread the sample on top of the sheets
of paper, taking care not to lose any soil off the paper or contaminating any adjacent samples.

       While spreading the sample to dry, pay particular attention to the sample code on the
DDRP Label A.  Where both the second and the third character of a routine sample code is the
digit "1", e.g., as in R11, do  not combine the sample with any other sample. Where either the
second or the third character of a routine sample code is a number other than "1", e.g., as in
R12 or R22, combine that sample  with those  routine samples having the same state, county,
pedon, and horizon code.  If a field duplicate sample has any designation other than "FDO", e.g.,
as in FD1 or FD2, combine  it with its companion field duplicates having the same state, county,
pedon, and horizon number.

       The combined samples are labeled  with the sample code designating the final bag of the
sample and the total number of bags of the  sample. For instance, if one had three samples
labeled R13	R23	and R33	all three samples would be combined on the drying table and
labeled R33 with the remainder of the sample code. Do not combine any samples that contain
a zero or an alphabetic character in the third space of the sample code.
                                          70

-------
       Disaggregate any large peds that impede the spreading of the sample over the entire
area of the paper.  Place an additional sheet of kraft paper loosely over the sample.  Attach the
sample's processing raw data form on the hook attached to the drying table. The original
canvas and plastic sampling bags should be kept on the floor beneath the sample as a second
check on sample identity.

NOTE: Immediately after a sample is spread to dry, field-moist pH is determined as outlined in
Section 3.2.

       Daily, the soil sample is to be stirred with gloved hands to facilitate drying. For the first
few days that a wet sample is spread, the bottom sheet of paper may need to be changed
daily in order to alleviate excessive moisture accumulation. Any observations of fungal or algal
growth should be noted on the bulk sample processing raw data form.

NOTE: Soils high in clay may harden nearly irreversibly if allowed to dry without  a preliminary
disaggregation of medium and coarse peds. An effort should  be made during the air-drying
procedure to disaggregate these peds while they are still somewhat moist or friable, before they
reach an air-dry state.

       Allow the sample to air dry until it is believed to be at or below the specified moisture
content.  This process generally takes about two days, although drying time may vary from one
day to a week or more.  At this time, an aliquot can be subsampled for the air-dry moisture
determination outlined in Section 3.4.
3.2 FIELD-MOIST pH DETERMINATION
3.2.1 General Considerations
                                           71

-------
       Immediately after a sample is spread to air dry, a field-moist subsample is collected and
pH in deionized water is determined.

       Soils high in salts, especially sodium, may interfere with the pH reading  and the
electrode response time.

       Clay particles may clog the KCI junction of the pH junction, slowing the electrode
response time.  Thoroughly rinse the electrode with deionized water between sample readings
to avoid this problem.

       Do not wipe the electrode dry with laboratory tissue or similar materials, and do not
remove the electrode from solution when the meter is not on standby, as either practice may
lead to polarization of the electrode.

       The pH can vary as much as one pH unit between the supernatant and soil sediment.
Always place the electrode junction at the same distance above the surface of the soil sediment
to maintain uniformity in pH readings.

3.2.2  Equipment

       Digital pH/mV meter, capable of measuring pH to 0.01 pH units, voltage potential to 1
              mV, and temperature to 0.5°C
       National Bureau of Standards (NBS)-traceable pH buffer sets of pH = 4.0  and pH = 7.0
              for calibration (two sets from different sources)
       NBS thermometer
       Combination pH and reference electrodes, high quality, low-sodium glass  [Geltype
              reference electrodes must not be used; an Orion Ross combination pH electrode
              or equivalent with a retractable sleeve is recommended]
                                            72

-------
        125-mL Nalgene bottles
        25-mL and 50-mL plastic containers
        Carboy, 13-gallon, filled with deionized water
        Field-moist pH raw data form
        Field-moist pH log book

3.2.3 Calibration and Standardization

        For storage and readings, the electrode need only be immersed to cover the liquid
junction of the reference electrode (typically about 2.5 cm).  Rinse the electrode with deionized
water between each sample and each buffer to prevent solution carryover.  Do not rub or blot
electrode dry because this may produce a static electric charge and thereby polarize the
electrode.

        To prepare the pH electrode  for use, move the sleeve covering the fill hole and fill the
reference reservoir to level of the hole  with 3M KCI solution.  Allow 5 minutes for the ceramic frit
to become wet with filling solution before immersing the electrode in the sample or the buffer
solution.  The retractable sleeve junction allows easy cleaning of clay particles  and insolubles
that clog the junction.

        Using the NBS pH 7.0 and pH 4.0 buffer solutions, calibrate the electrode as specified in
the electrode manual for a two-buffer calibration.  Place about 35mL of pH 7.0  buffer solution
into a 50-mL plastic container, stir, wait 30 seconds, and take the reading.  Record in the pH log
book the temperature of the  solution on the meter and also the reading on the  NBS
thermometer.  Adjust the temperature control on the meter, read the  pH after equilibration,  and
adjust the meter if necessary. Perform the steps again using the pH 4.0 buffer and the slope
control on the meter.  Repeat measurements and adjustments until readings for both buffer
solutions are within 0.05 pH unit of the respective  known values.
                                            73

-------
       Calibrate the pH meter before each run of sample analyses.  All readings and
adjustments should be recorded in the pH log book and initialed by the analyst.

3.2.4  Quality Control

       Before analyzing the samples, measure a quality control check sample (QCCS) of pH 4.0
using a different batch of NBS stock buffer solution than  that used for the meter calibration.
Subsequently, after every 10 samples and after the final sample of the run, measure another
QCCS of pH 4.0.  The field-moist pH raw data form (see Appendix A) contains an entry field for
these data indicated by "QCCS". The value  recorded for each QCCS should be 4.00 ± 0.05.  If a
QCCS value is not within that range, recalibrate the electrode according to the procedure
described in Section 3.2.3 above.  If acceptable results still cannot be obtained, check:  (1) that
the reference junction of the electrode is clean,  (2) that the wiring straps into the meter are
firmly connected, (3) for static discharge from another instrument or from the operator, and (4)
to ensure that sufficient filling solution is contained within the electrode.  If the problem
persists, replace the electrode, the meter, or both.

       Within each run of pH analysis, an internal duplicate sample of a field-moist soil is to be
obtained and analyzed for every 10th sample. The letter "D" on the raw data form indicates
when a duplicate is to be subsampled.  The "D" replaces  the first character of the sample code
for the sample that was selected.  The duplicate sample  and its corresponding routine sample
pH values should have a standard deviation for the pair of 0.15 pH units or less.  If the
standard deviation is higher than this value, the data should be reviewed by the laboratory
manager and the QA manager. A decision  may be made  to re-run the pH analysis on these
samples.
                                            74

-------
       Also within each run, a manager's sample (MS) is included.  The MS is an audit sample
having characterized audit windows, but its concentration range is unknown to the analyst
performing the pH analysis.  The sample is used to assess accuracy of the run.

       If a run of pH analysis consists of fewer than 10 samples, the QCCS and internal
duplicate samples should be placed as the final samples in the run.

NOTE: Because the pH is measured immediately after a sample is spread to air dry, the
number of samples in each run varies.  It is necessary to adhere  to all calibration protocols and
QC specifications for all runs, regardless of size.

3.2.5  Procedure

       After spreading and stirring the field-moist soil, scoop approximately 50g  of mineral soil
or 15g of organic soil into a 125-mL Nalgene bottle and record the sample code on the label of
the bottle.  Refrigerate the bottles at 4°C until such time as the analyst is ready to begin a run.
NOTE: Some soils may be in a saturated state when received. Follow the procedure as stated;
however, note the saturated condition of the soil on the bulk sample processing raw data form.

       Place approximately 20g of mineral soil or 5g of organic soil into a pre-numbered 25-mL
plastic container.  Add 20ml_ of deionized water for mineral soils or 25mL for organic soils.
Record the sample code and the container number on the field-moist pH raw data form.

       Allow the sample to absorb the solution for approximately 60 seconds, then thoroughly
stir the soil-solution mixture for 10 seconds  with a glass  stirring rod. Rinse the rod with
                                           75

-------
deionized water after stirring each sample.  Stir again for 10 seconds after 15, 30, 45, and 60
minutes. Allow the suspension to settle for 30 minutes.

       Place the pH electrode  in the supernatant of the soil suspension.  For mineral soils, the
electrode junction is to be below the solution surface and above the soil-solution interface.
Although organic soils normally absorb all of the free water available, an acceptable reading
generally is obtained if the electrode junction is below the meniscus of the organic material.

       Record the pH of each  sample (PH_MP) on the field-moist pH raw data form.  Once the
pH analysis is completed for a run of soils, enter the data into the computer entry and
verification system. If there are no flags generated (see Appendix B), transcribe the pH values
onto the appropriate entry field of each bulk sample processing raw data form.  Any sample
with a reading of  pH 6.5 or higher is to be reported to the QA manager immediately.

       After the pH measurements are completed, store the electrode in storage solution.  Do
not allow the sensing element  or reference junction to dry out.  The level of the storage solution
should be one inch below  the filling solution level. The electrode sleeve is to be closed during
storage. Check periodically to ensure that the electrode reservoir is full of storage solution.
3.3 ORGANIC MATTER DETERMINATION BY LOSS-ON-IGNITION

3.3.1 General Considerations

       Loss-on-ignition (LOI) is the method used to determine percent organic matter of the
MASS samples.  Because organic samples are oxidized at high temperatures, percent organic
matter can be calculated on a weight-loss basis.  From the percent organic matter, the percent
                                            76

-------
organic carbon can be estimated. A modified version of the method described in MacDonald
(1977) is used.

       Laboratory personnel should use extra caution when working around the  muffle furnace,
as temperatures of 450°C are common.  The furnace should be activated only in  an operable
fume hood.

       Crucibles must be thoroughly cleaned after each use, as the crucible weights are pre-
determined and entered into the computer verification system.  Tongs or finger cots should be
used when transferring crucibles from one location to another.

3.3.2 Equipment

       Open-pan balance, accurate  to O.Otg
       Small crucibles, pre-numbered
       Tongs
       Finger cots
       Desiccator, with desiccant and crucible plate
       Muffle furnace in operating fume hood
       Furnace gloves
       Convection oven
       Loss-on-ignition raw data forms
       Loss-on-ignition log book

3.3.3 Quality Control

       Within each run of LOI analysis,  an internal duplicate sample of a routine soil sample is
to be obtained and analyzed. The letter "D" replaces the first character of the sample code for
                                            77

-------
the sample that was selected.  The duplicate sample and its corresponding routine sample LOI
values should have a relative standard deviation for the pair of 15 percent or less.  If the
relative standard deviation is higher than this value, the data should be reviewed by the
laboratory manager and the QA manager. A decision may be made to re-run the LOI analysis
on this sample.

       Also within each run, a  manager's sample (MS) is included.  The MS is an audit sample
having characterized audit windows, but its concentration range is unknown to the analyst
performing the LOI analysis. The sample is used to assess  accuracy of the run.

       If a run of LOI analysis consists of fewer than 10 samples, the internal duplicate and
MS should be placed as the final samples in the run.

3.3.4  Procedure

       The LOI analysis can be performed at any time while the samples are air drying. The
data obtained  are to be recorded on the loss-on-ignition raw data form  (see Appendix A).

       Fill a small pre-numbered and pre-weighed crucible about two-thirds full of soil.  Record
the sample code and crucible number on the raw data form under "CRUC_NO". Oven dry the
crucible plus sample overnight, at 105°C for mineral samples or 60°C for organic samples.

       The following morning,  remove the crucible from the convection oven and place it in a
desiccator for  30 minutes.  Weigh the crucible plus sample and record its oven-dry weight  under
"OD WP.
                                           78

-------
       Carefully place the crucible plus sample in a muffle furnace and secure the door.  Set
the furnace temperature at 450°C and turn on the unit. The sample should be left overnight at
this temperature.

       The following morning, turn the furnace off and allow to cool for 30 minutes. Carefully
remove the crucible from the furnace and allow to cool in the desiccator with a dish of
desiccant.

NOTE: Because a vacuum may be created when cooling extremely hot samples, extra caution
should be exercised when opening the desiccator. If a vacuum has formed, the lid  may be
extremely hard to remove and/or a rush of air may enter the desiccator upon opening.

       Finally, weigh the crucible plus  remaining soil and record this weight under "ASHED_WT"
on the raw data form.  Enter the data  into the computer entry and verification system.  If there
are no flags generated, the  sample can undergo further processing.

       If the calculated organic matter content is 20 percent or more,  the sample is assumed
to be an organic sample. If any samples which were  initially labeled as mineral are determined
to have 20 percent or more  organic matter, batch these with the organic samples, or vice versa.
Report these samples to the QA manager and the soil sampling task leader as soon as
possible.

3.3.5  Calculations

       The following calculation is performed by the computer  after the raw data are entered,
but it is presented here for internal use.  It is advantageous to manually check a few samples
using the calculation  in order to understand the procedure and  to test  the accuracy of the
computer program.
                                           79

-------
       Percent Organic Matter (OMJ.OI) =   OD WT - ASHED WT
                                           —	—   x100
                                            OD WT - CRUC WT
3.4 AIR-DRY MOISTURE DETERMINATION

3.4.1 General Considerations

       To ensure that each sample is at an acceptable moisture level for further processing, it
is necessary to perform a moisture analysis for each soil sample.  The procedure for assessing
air-dry  moisture is not to be performed, however, until the soil is believed to be air dry.

3.4.2 Equipment

       Aluminum weighing dishes, pre-numbered
       Forceps or finger cots
       Open-pan balance, accurate to 0.01g
       Convection oven
       Desiccator and desiccant
       Air-dry moisture raw data forms
       Air-dry moisture log book

3.4.3 Quality Control

       Within each run of moisture analysis, an internal duplicate sample of a routine soil
sample is to be obtained and analyzed. The letter "D" replaces the first character of the sample
code for the sample that was selected. The duplicate sample and its corresponding routine
sample moisture values should have a  relative standard deviation for the pair of 15 percent or
                                           80

-------
less.  If the relative standard deviation is higher than this value, the data should be reviewed by
the laboratory manager and the QA manager.  A decision may be made to re-run the LOI
analysis on this sample.

       Also within each run, a manager's sample (MS) is included. The MS is an audit sample
having characterized audit windows, but its concentration range is unknown to the analyst
performing the moisture analysis.  The sample is used to assess accuracy of the run.

       If a run of moisture analysis consists of fewer than 10 samples, the internal duplicate
and MS should be placed as the final samples in the run.

3.4.4  Procedure

       Data for this procedure are recorded on the air-dry moisture raw data form (see
Appendix A).  Once a soil sample has been determined to be air dry, the information is also
entered on the bulk sample processing raw data form.

       Thoroughly mix the air-dry sample with gloved  hands.  Transfer a subsample of
approximately 15g into a pre-numbered aluminum weighing dish of known weight. Because the
aluminum  weighing dishes are manufactured to be a nearly-constant weight, the average weight
of ten labeled aluminum weighing dishes may be used.  Enter the dish number on the air-dry
moisture raw data form under "TIN_NO". Handle the weighing dish with forceps or finger cots.

       Weigh the dish plus sample to the nearest 0.01g and record this initial weight under
"INIT_WT". Place the  dish in a convection oven which has equilibrated at 105°C for mineral
samples or 60°C for organic samples.  Allow the sample to oven dry overnight at this
temperature.
                                           81

-------
       The following morning, remove the sample from the oven and allow to cool for 30
minutes in a desiccator.  Weigh the dish plus sample to the nearest 0.01g and record this oven-
dry weight under "OD_WT".

       Enter the data from the raw data form into the computer entry and verification system
and check for any flags (see Appendix B). If the calculated moisture content for a mineral soil
is above 2.5 percent for mineral soils or above 6.0 for organic soils, allow the bulk sample to
continue air-drying and repeat the procedure at a later time.  However, if the sample  is below
the cutoff percent moisture content for the soil type, then the air-dry sample may be  rebagged
and placed in cold storage or it may immediately undergo the next stage of processing.

3.4.5  Calculations

       The following calculations are performed by the computer after the raw data  are entered,
but they are presented here for internal  use.  It is advantageous to manually check a few
samples using these calculations in order to understand the  procedure and to test the accuracy
of the computer program.
       Percent air-dry moisture (MOIST P)  =   INIT WT - OD WT
                                            	  xfOO
                                            OD WT - TIN WT
3.5 DISAGGREGATION AND SIEVING

3.5.1  General Considerations

       After a bulk soil sample has been determined to be air dry, it is ready to be
disaggregated and sieved in order to remove rock fragments and to prepare the sample for

                                           82

-------
homogenization and subsampling.  The disaggregation and sieving areas should be covered
with protective layers of kraft paper and cleaned after each sample has been sieved.  Dust
masks and protective clothing should be worn at all times. Some form of ventilation and
heating should be provided for the  sieving area.

       A source of compressed air is necessary to clean the surfaces of the equipment, e.g.,
rolling pin, sieves, and gloves, after processing each sample.  However, because water and oil
tend to accumulate in the tank and conduit, the air must be passed through a trap to collect
the offending vapor and liquid.

       In addition to cleaning with compressed air, the surfaces of the processing equipment
may be wiped with clean Kim wipes® if soil is adhering to the rolling pin or to the frame of the
sieve.  Used tissues are to be discarded after each sample is processed in order to avoid
contamination.

3.5.2 Equipment

       Rolling pin, wooden
       Rubber stoppers
       Brass sieves, square-holed: 2-mm and 4.75-mm
       Clean sieving table and work area
       Kraft paper, 36" wide rolls
       Air compressor, with in-line  filter
       Gloves, unpowdered
       Open-pan balance,  accurate to 0.1g
       2-L Nalgene bottles
       DORP Label A rolls
       Wide-mouth funnel
                                           83

-------
       Small cellophane bags with twist ties
       Dust masks
       Bulk sample processing raw data forms

3.5.3  Procedure

       A bulk sample must be weighed in  its  entirety before it is disaggregated and sieved in
order  to properly calculate the percentage  of rock fragments in the sample. Weigh the bulk
sample while still in its  plastic bag to the nearest 0.1g and record this weight under "TOTAL
BULK_WT" on the bulk sample processing  raw data form.  Large pieces of undecomposed
stems, moss, twigs, or  roots are to be removed from a mineral sample before its bulk weight is
determined.

       Place a 1-m by 1-m sheet of kraft paper on top of the sieving table.  Next, place a 60-cm
by 60-cm sheet of kraft paper on top of the table and spread small portions of the bulk sample
within the confines of the sheet. Carefully examine the nature of the rock fragments within the
sample and determine the amount of pressure to apply to the sample in order to disaggregate
the soil peds without fracturing or crushing the fragments.

       Place another 60-cm by 60-cm layer of kraft paper over the sample and gently roll across
the sample with a wooden rolling pin; enough force should be applied to break up the peds, but
not so much that weathered rock fragments are  crushed. Place a portion of sample in the 2-
mm mesh sieve and gently push the soil through the  sieve with a rubber stopper. Attempt to
include any soil adhering to rock fragments.  Remove and discard any small remaining roots
that nest on the sieve.  Continue sieving until  all  of the  less than 2- mm soil has passed
through the 2-mm sieve.
                                           84

-------
       Save the less than 2-mm sample in a 2-L Nalgene bottle pre-labeled with a DDRP Label
A and place the bottle in cold storage. Continue processing the remaining rock fragments as
described in Section 3.6 below.
3.6 ROCK FRAGMENT DETERMINATION

3.6.1  General Considerations

       The considerations for this procedure are the same as those outlined for disaggregation
and sieving in Section 3.5.1.

3.6.2  Equipment

       As outlined in Section 3.5.2

3.6.3  Procedure

       Save the rock fragments which could not pass the 2-mm sieve and place them on the
4.75- mm sieve. Retain in separate cellophane bags the material which passes the 4.75-mm
sieve and that  which does not pass.

       Weigh to the nearest 0.1g  the fragments which passed the 4.75-mm sieve but were
retained by the 2-mm sieve.  Record this weight on the bulk sample processing raw data form
under "2- TO 4.75-MM ROCK FRAGMENTS".  Then weigh the fragments that were retained by the
4.75-mm sieve.  Record this weight under "4.75- TO 20-MM ROCK FRAGMENTS".
                                          85

-------
       Place the rock fragment bags for each sample in the leftover plastic bag from the bulk
sample. Make sure that a DORP Label A with the appropriate sample code and set ID is
affixed to the bag.  Save all rock fragment bags, organized by set ID, until instructed by the soil
sampling task leader to discard or transfer the fragments to another location.  It is not
necessary to keep the rock fragments in cold storage.

3.6.4  Calculations

       The following calculations are performed by the computer after the raw data are entered,
but they are presented here for internal use.  It is advantageous to manually check a few
samples using these calculations in order to understand the procedure and to test the accuracy
of the computer program. The percentages of fine gravel (RF_FG) and medium  gravel (RF_MG)
are calculated as follows:

       RF FG =  2- TO 4.75-MM ROCK FRAGMENTS
                 	 x100
                         TOTAL BULK_WT

       RF MG  =  4.75- TO 20-MM ROCK FRAGMENTS
                                                   x100
                          TOTAL BULK WT
3.7 HOMOGENIZATION AND SUBSAMPLING

3.7.1 General Considerations

       In order to obtain representative volumes of soil suitable for detailed analysis at the
analytical laboratories, it is necessary to prepare homogenous subsamples from the less than
2-mm soil fraction through the use of a riffle splitter.
                                          86

-------
       The riffle splitter is to be cleaned with compressed air after each sample is split. If the
operator suspects that the riffle splitter is still not thoroughly clean, the riffle splitter should be
washed with deionized water and allowed to air dry completely. The work area is to be cleaned
with a brush or hand vacuum after each sample is processed. The work area should be vented
with an operating fume hood.

       The preparation laboratory prepares two different subsamples because different
analytical laboratories are responsible for the general analyses of 47 physical and chemical
parameters and for the elemental analysis of total carbon, nitrogen, and sulfur. The general
analytical sample should weigh approximately 500g, while the elemental analytical  sample
should weigh approximately 50g.  Each analytical sample is derived through homogenization and
subsampling by the procedure described below.

3.7.2  Equipment

       Riffle  splitter, Jones-type, closed-bin, two-pan, 1.25-cm baffles
       Distribution pan
       Open-pan balance, accurate to 0.1g
       Nalgene bottles:  2-L, 500-mL,  125-mL
       DDRP Label A rolls
       Wide-mouth funnel
       Oust mask
       Gloves, unpowdered
       Air compressor, with in-line filter

3.7.3  Procedure
                                            87

-------
       For each sample, pre-label one 500-mL and one 125-mL Nalgene bottle with the DORP
Label A information. Position the two receiving pans under the riffle splitter.  Pour all of the
less than 2-mm soil from the 2- L Nalgene sample bottle evenly across the baffles of the riffle
splitter. Transfer the soil from  each receiving pan into the distribution pan and replace the
receiving pans under the riffle splitter.  Repeat this step five times in succession with the
material in each receiving pan.

       Next, pour the sample evenly across the baffles and place the soil from one receiving
pan into the 2-L bottle.  Transfer the soil in the other receiving pan to the distribution pan and
continue splitting as necessary until approximately 500g of soil occupies one of the receiving
pans.  Place the entire contents of the pan into the pre-labeled 500-mL bottle (general analytical
sample). Repeat the splitting using the remaining soil until approximately 50g of soil occupies
one of the  receiving pans; transfer the entire contents of the pan into the pre-labeled 125-mL
bottle (elemental  analytical sample).

       Transfer all  leftover soil into the  2-L bottle which wilt serve as a DORP-MASS archive
sample.  The analytical samples are grouped by set ID and placed in cold storage until they are
batched and shipped via overnight express service to the analytical laboratories. The archive
samples are grouped by batch  ID and remain in cold storage indefinitely.
3.8 BULK DENSITY DETERMINATION

       Density is defined as weight per unit volume expressed in units of g/cm3.  The bulk
density of a soil is defined as the weight of dry soil per unit volume including the pore space.
For mineral soils, bulk density generally ranges from 0.6 to 2.0 g/cm3.  With increasing organic
matter content, soils generally exhibit a decrease in bulk density because organic matter has
higher porosity and lower density than mineral particles of the same diameter.
                                            88

-------
       As in past DORP surveys, the clod method is the primary method for determining bulk
density in the MASS.  Two alternate methods are also being attempted for soil horizons that fail
to yield satisfactory clods; one is a volume replacement (VR) method which utilizes a known
volume of small foam beads packed into a cylinder to replace a selected volume of soil
excavated from a given horizon, while the other is a volume filling (VF) method that is used if
the clod or VR methods do not produce  representative samples.

3.8.1  Clod Method

3.8.1.1 General Considerations -

       Where possible, three replicate clod samples are extracted from each horizon.  The
average bulk density of the replicates is  assumed to be the bulk density of that particular
horizon.  Analysis of  the clods is based on the method described in USDA/SCS (1984).

       Laboratory personnel should use caution when working around the muffle furnace, as
temperatures of up to 450°C are common.  The furnace should be activated only in an operable
fume  hood.

       Evaporating dishes must be thoroughly cleaned after each use, as their weights are pre-
determined and entered into the computer verification system.  Tongs or finger cots should  be
used  when transferring the dishes from one location to another.

       Extra care must be taken in the use of the Saran® powder and acetone solution (see
Section 3.8.1.3). When mixed, the resulting solution has a tendency to volatilize hydrogen
chloride gas which can cause deleterious health effects.  The solution should be  used only in an
operating vented fume hood.  A cartridge-type organic vapor mask and laboratory coat  must be
worn.
                                           89

-------
3.8.1.2 Equipment -

       Saran:acetone clod dipping mixture, approximately 1:5 by weight
       Carboy, 13-gallon, filled with deionized water
       Organic vapor mask, cartridge-type
       Open-pan balance, accurate to 0.1g
       Open-pan balance, accurate to 0.01g
       Beakers, 1000-mL and 2000-mL
       Ring stand
       Test tube clip, adjustable
       Convection oven
       Evaporating dishes,  pre-numbered, tolerance to 450°C
       Furnace gloves
       Tongs
       Muffle furnace
       Desiccator and desiccant
       Sieve, square-holed, 2-mm
       Paper bags, pre-labeled
       Clod bulk density raw data form

3.8.1.3 Mixing  the Clod Dipping Solution -

NOTE: The clod dipping solution must be mixed only under an operating vented fume hood.  The
operator must wear a cartridge-type organic vapor  mask and laboratory coat while mixing the
dipping solution.

       The Saran® powder is packaged in 540g increments by EMSL-LV QA staff.  One packet
mixed with approximately 3400ml_ of acetone provides a nearly 1:5 by weight mixture resembling
                                            90

-------
light syrup.  One-gallon paint containers are used for mixing and storing the solution. Slowly
add the contents of each packet to the acetone while stirring.  Continue stirring to ensure
thorough dissolution of the powder.  The solution normally is well  mixed when it has the color
and consistency of amber syrup and there is no gummy saran residue on the edges or bottom
of the container.  If the solution is a milky color,  additional mixing is required. The powder has
a tendency to precipitate and clump around a container's lower rim when  mixed with acetone.
To prevent this, use a long stir bar and thoroughly scrape the inner rims.  If possible, use a
magnetic stirrer and a stir bar to mix the ingredients in a beaker until well blended.

       After mixing the solution, the container should be capped as tightly as possible to
prevent leakage of the solution or evaporation of the acetone.

3.8.1.4 Procedure -

       All raw data obtained from  the clod analysis are recorded on the clod bulk density raw
data form (see Appendix A).  For each clod to be analyzed, record the set ID, sample code, and
the replicate number on the form.  Examine the label attached to the clod  and record the
number of field dips performed by the sampling crew under "FIELD_DP" on the form.  Weigh the
clod, without removing the label, to the nearest 0.1g and record this weight under
"FIELD_WT/MOIST".

       Suspend the clod in a convection oven and dry overnight at 105°C for  mineral samples or
60°C for organic  samples.  The following morning, transfer the  clod into a  desiccator for 30
minutes. Weigh  the sample to the nearest  0.1g and record this weight under "FIELD_WT/DRY'.

       Hang the clod on a suspended curtain rod within an operating vented fume hood.  Dip
the clod into a container of 1:5 by weight saran:acetone mixture (see Section 3.8.1.3) for three
                                            91

-------
seconds and then allow the coating to dry for approximately 15 minutes.  Do not allow the
mixture to impregnate that portion of  the hairnet hanging above the top of a clod.

       Apply additional dips as necessary until the clod does not produce excessive bubbles
and a coating of saran is obtained that appears to be impervious to water.  Usually three to six
dips is sufficient for most clods. Record the total number of dips made in the laboratory under
"LABJDP",  then reweigh the clod to the nearest 0.1g and record this weight under "LAB_WT'.
Also, record the dry weight of a 500- g Class P balance weight immediately before and after a
run of samples.  This  weight is used to calibrate the density readings and is charted on  a daily
basis.

       Add approximately 1600-mL deionized water to a 2000-mL beaker. Place the beaker of
water on a balance and tare. Record the temperature of the water under "TEMP" so that the
density of the water can be calculated.  Also, record the submerged weight of a 500-g Class P
balance weight immediately before and  after a run of samples.

       Suspend the clod over the beaker by attaching the top of the hairnet to the test tube
clip, then lower the clod gently into the  water until the top of the clod is entirely submerged.
Promptly, after three seconds,  record  the weight displayed on the balance to the nearest 0.1g
under "CLOD_H2O".  If the clod floats, forcibly submerge it by pushing down with forceps and
record the weight.  Note on the raw data form, with  a "V under "FLOAT1, that the clod floated.
Re- tare the balance before proceeding  with other clods.

NOTE: When submerging, ensure that the clod is not touching the edge of the beaker and that
the clod label hangs freely.  Occasionally, air bubbles may rise from the clod and the weight
reading on the balance does not stabilize but steadily decreases.  This occasionally occurs  in
clods sampled from relatively porous  surface horizons and simply indicates that all of the
primary macropores along the  exterior of the clod were  not thoroughly water-sealed.  By reading
                                            92

-------
the weight immediately after submersion, error from this source of instability is minimized. Note
the bubbling on the raw data form under "COMMENTS". If the data for this clod do not pass
the verification, a "Y" can be entered under "PROB" that will direct the computer program to
delete that value from the average of the replicates. Also note any other problem, such as a
broken clod, under "COMMENTS".

       After submersion, place the clod  in a pre-numbered evaporating dish and record the dish
number on the raw data form under "CRUCJMO".  Remove as much of the clod label and hairnet
as possible and place in a large plastic bag with others from the same run.

       Place the evaporating dish plus clod in a muffle furnace equilibrated at  400°C.  The
muffle furnace must be within an operating vented fume hood. Allow the saran coating to be
burned off the clod surface for two hours. Allow the furnace to cool, then carefully remove the
evaporating dish from the muffle furnace.  After the clod has cooled thoroughly, place  it in a
pre-labeled paper bag.

       Disaggregate the clod through a  2-mm sieve. Weigh any rock fragments retained on the
sieve to the nearest 0.01g and record this weight under "R_FRAG".  Place the rock fragments
into the paper bag and archive all such bags.

3.8.1.5  Assumptions -

       It is assumed that the weight of  each two-second saran  coating applied in the field is
equal to the weight of each three-second coating applied in the laboratory, as some of the
dipping solution normally is absorbed into the clod when it is applied in the field.
                                           93

-------
       It is assumed that the specific gravity of air-dry saran is 1.30 g/cm3 and that the coating
loses 15 percent of its weight upon oven-drying. It is assumed that the particle density of the
rock fragments is 2.47 g/cm3.

       It is assumed that the average gross weight of the hairnet and clod label is 1.88g and
the net weight of the submerged portion of the hairnet is 0.20g.

3.8.1.6 Calculations -

       The following calculations are performed by the computer after the raw bulk density
data are entered.  The calculations are presented here for internal use.  It is advantageous to
manually check a few samples using these calculations in order to understand the  procedure
and to test the accuracy of the computer program. A table of temperature- corrected water
density values is provided in Appendix C.

       Use the following calculations to derive the weights and volume of the air-dry and oven-
dry saran coatings:

       SARAN_WT (g)  =  {[(FIELD_DP + LAB_DP] x [LAB_WT - FIELD_WT/DRY]}
                                               LABJDP
       OD SARAN_WT (g)  =   SARAN_WT x 0.85
       SARAN VOL (cm3)  =  SARAN_WT
                               1.30

       Use the following calculations to derive the volumes of rock fragments, water
displacement, and soil/pore fraction:

       R_FRAG VOL (cm3) =  R_FRAG
                               2.47
                                           94

-------
       WATER VOL (cm3) =      CLOD_H2O
                            DENSITY OF WATER
       SOIL/PORES VOL (cm3) =  [WATER VOL - (R_FRAG VOL + SARAN VOL)]

       Finally, use the following calculation to derive the oven-dry bulk density (BD_CLD) for the
individual clod:

       BD_CLD  (g/cm3)  =   [LAB_WT - (R_FRAG + OD SARAN_WT +  1.88)]
                                    SOIL/PORES VOL - 0.20

3.8.2 Known Volume Methods

3.8.2.1  General Considerations -

       A volume replacement sample is designated by a "VR1" or "VR2" as the first three
characters of the sample code.  The samples are subject to variable volume estimates which
are recorded on the SCS-SOI-232 field data form and transcribed into the sample receipt log
book. Subtracting the initial from the final volume yields the estimated volume of sample
collected.

       A volume filling sample is designated by a "VF1" or "VF2" as the first three characters of
the sample code.  The volume of this type of sample is based on the absolute height of a 250-
mL beaker, which is a constant  300cm3.

       The known volume samples are processed in a manner similar to the method described
in Blake (1965).

3.8.2.2  Equipment -
                                          95

-------
       Small paper bags, pre-weighed
       Open-pan balance, accurate to 0.01g
       Convection oven
       Brass sieve, square-holed, 2-mm
       Known volume bulk density raw data forms
       Known volume log book
       Volume data from sample receipt log book

3.8.2.3  Procedure -

       Known volume bulk density samples are shipped from the field in small plastic bags.
Upon passing the sample receipt verification, these samples can be placed out to air dry in the
drying area.  The drying facilitates the transfer of the entire sample from its plastic bag into a
pre- labeled paper bag.  Do not remove any material from the bag, even if the material is
thought to be morphologically different from the type of horizon sampled.

       Record the set ID and sample code of the VR or VF sample on the known volume bulk
density raw data form (see Appendix A).  Weigh the paper bag plus sample to the nearest 0.01g
and record this weight under "AIR_WT".  Place the bag in a convection oven equilibrated at
105°C for mineral soils and 60°C for organic soils. Allow the sample to oven dry overnight.

       The following morning, remove the sample from the oven and allow to cool for 30
minutes. Weigh the sample to the nearest 0.01g and record this weight under "OD_WT". Sieve
the contents of the bag through a 2-mm sieve to separate any rock fragments from the fines.
Weigh the rock fragments to the nearest 0.01g and record this weight under "R_FRAG".  Place
the rock fragments in the paper  bag and archive all such bags of fragments.
                                           96

-------
       Transcribe the computed volume (V, minus V() for a VR sample from the sample receipt
log book to the "VOL" column on the raw data form. For a VF sample, 300 cm3 can be
automatically entered under "VOL".

3.8.2.4 Assumptions -

       The particle density for rock fragments in a VR bulk density sample is derived individually
by a submersion technique similar to that used in the clod analysis. It is assumed that the
particle density for rock fragments in a VF bulk density sample is based  on four classes defined
by different soil parent materials.

3.8.2.5 Calculations -

       The following calculations for known volume bulk density (BD_KV) are performed by the
computer entry and verification system after the raw data are entered. It is advantageous to
manually check a few samples using these calculations  in order to  understand the procedure
and to test the computer program.

       FINES_WT (g)  =  OD_WT - R_FRAG
       BD_KV(g/cm3)  =            FINES_WT
                         VOL - (R_FRAG / particle density)
                                          97

-------
                                      REFERENCES
Blake, G. R. 1965. In  Black, C. A. (ed.) Methods of Soil Analysis, Part 1.  American Society of
Agronomy,  Madison, Wisconsin.

MacDonald, C. C. 1977.  Methods of Soil and Tissue Analysis in the Analytical Laboratory.
Maritimes Forest Research Centre Information Report, M-X-78.  Fredericton, New Brunswick,
Canada.

Papp, M. L, R. D. Van Remortel, C. J. Palmer, G. E. Byers, B. A. Schumacher, R. L. Slagle, J. V.
Burton, K. C. Shines, J. E. Teberg, and M. J. Miah.  1989. Direct/Delayed Response Project:
Quality Assurance Plan for Soil Preparation and Analysis in the Mid-Appalachian Region of the
United States.  EPA/600/4-89/031.  U.S.  Environmental Protection Agency, Las Vegas, Nevada.

USDA/SCS.  1984. Soil Survey Laboratory Methods and Procedures for Collecting Soil Samples.
Soil Survey Investigations Report No. 1.  U.S. Government Printing Office, Washington, D.C.

U.S. Environmental Protection Agency.  1985a. Direct/Delayed Response Project: Soil Survey
Data Quality Objectives  (Draft). U.S. Environmental Protection Agency, Washington,  D.C.

U.S. Environmental Protection Agency.  1985b. Direct/Delayed Response Project, Long-term
Response of Surface Waters to Acidic Deposition: Factors Affecting Response and a Plan for
Classifying that  Response on a Regional Scale.  U.S. Environmental Protection Agency, Corvallis,
Oregon.
                                            98

-------
                               Appendix B

                Preparation Laboratory Data  Forms
      The folbwing labels and forms were used by the preparation laboratory to label samples,
enter raw data, and calculated final data for the Mid-Appalachian Soil Survey.
                                    99

-------
                     DDRP Label A
Date Sampled:	
            D  D M M M Y Y
Crew ID:	


Site ID:	


Sample Code:	—	


Horizon:	                 Depth:	—	cm.
Set I.D.:.
                    DDRP Label B
                Batch ID:
                Sample No.:	
                      100

-------
SOIL SAMPLE RECEIPT FORM
SAX CODE





















CREW
ID





















SITE
ID





















SET
ID





















DATE
SAMPLED





















DATE
SHIPPED





















DATE
RECEIVED





















RECEIVED
BT





















SAMPLE CONDITION
WET/DRY (H/D) BAG SPLIT (B/S)
SIEVED/USSIEVED (S/D)
UNDER VOLUME (UV)





















CLOD
NUMBER





















KNOWN
VOLUME
CAVITI
SAMPLE
NUMBER





















COMMENTS























-------
SAMPLE ID:




  SITE ID:




   SET ID:




 BATCH ID:
  BULK SAMPLE RAW  DATA




  	    DATE SAMPLED:




 	             DATE REC'D:




 	          PROCESS START:




            PROCESS COMPLETE:
SOIL TYPE:  M/0



 Initials:
                    FIELD pH:



                    Initials:
                       AIR SAMPLE DRYING:




                   Date      % Moisture    Initials
 Date:
                          TOTAL BULK WT:
                         Initials:
 2 to 4.75 mm:




       Date:
  ROCK  FRAGMENT WT:




	.	g      4.75 to  20 mm:




 /	          Initials:
ENTERED IN COMPUTER:  Date:
                          Initials:
COMMENTS:
                               102

-------
pH RAW DATA
DATE    /     /
SAMPLE CODE
Init OCCS










D
QCCS










D
SAHP.
*
























pH
























SAMPLE CODE
QCC5










D
QCCS










Final QCCS
SAMP.
*
























PH
























    103

-------
                                          SAMPLE LOSS ON IGNITION DATA
ALL WEIGHTS REPORTED IN 0.01 GRAMS
  SAM CODE
                             CROC.
                              NO.
AIR
ST.
0. D.
 wr.
ASHED
 HT.
                                                             DATE
                                                                    ID
                                                                          COMMENTS
                                                        104

-------
                                     SOIL SAMPLE AIR DRT DETERMINATION
ALL WEIGHTS REPORTED TO 0.01 GRAMS
  SAM CODE
                              TIN
                              NO.
TIN
WT.
 AIR DRY
(TIN+SOIL)
 OVEN DRY
(TIN+SOI1)
                                                                    DATE
                                                                             ID
                                                                                   COMMENTS
                                                       105

-------
                                                                                 CLOD  BULK DENSITY RAH DATA
                                                                                                                                                                                page 	/_
                       ALL HEIGHTS RECORDED TO 0.01 GRAMS
o
o>
                                           REP
                                                FID DP
                                                         moist/dry
                                                                                        CLOD H20
                                                                                                                                                    T/H

-------
                                           10)0101 VOLUXE BULK DENSITY DATA
                                                                                                 PAGE
ALL WEIGHTS RECORDED TO 0.0  GRANS
  SAM CODE
                      AIR W    00
                                         R FRAG
                                                   VOL
                                                  (CC.)
                                                           DATS
                                                                     ID
                                                                            COMMENTS
                                                      107
                                                              \

-------
DIRECT/DELAYED RESPONSE  PROJECT (DDRP)
                 FORM 101
                                       Date received
                                       by Data Mgt.
                                                     D D M M M Y Y

Batch ID
Crew ID
Prep Lab ID

Lab Set Sent to

Set ID
Date Samj
Date Reef
Date Prej
No. of Se
Sample
No.
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Date Shipped



imples _ _
Site ID











































Sample Code













































	



	


Mineral or Organic Samples 	
Set
ID











































Signature of Preparation Laboratory Manager
Comments :
Coal
Fragn
2- 4
4.7










































se
tents
4.7-
20.










































Moisture
%
Field










































Air Dry












































PH










































LOI










































Bulk
Density
g/cm3











































                     108

-------
DIRECT/DELAYED RESPONSE PROJECT (DDRP)
           SHIPPING FORM 102
                            DATE RECEIVED

                            BY DATA MGT.	
                                       D  D M  M  M  Y Y
0 D M M M Y Y
Prep Lab ID Dale Received
Batrh ID Dale Shiooed
Analytical Lab
SAMPLE NO
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
ID

BULK SAMPLES
SHIPPED








































RECEIVED








































Signature of Preparation Laboratory Manager
Comments
PH <: 85








































	
RETURNED TO LEMSCO
(CHECK V IF YES)










































          H COPY TO SMO
                     M COPY TO EMSLIV
                   109

-------
                                Appendix C

                  Preparation Laboratory Computer
               Entry/Tracking/Verification Procedures
      The following text provides an overview of the computer tracking and verification program
designed for use at the preparation laboratory during the Direct/Delayed Response Project Mid-
Appalachian Soil Survey. The system is designed to track the progress of samples, calculate final
data values, perform verification checks, and assign flags to any detected aberrant values.
                                      110

-------
    PREPARATION LABORATORY COMPUTER ENTRY/TRACKING/VERIFICATION PROCEDURES


     The  following  text provides  an overview  of the  computer  tracking and
verification program designed for use at the preparation laboratory during the
DDRP Mid-Appalachian soil survey.   The system  is designed to track the progress
of samples,  calculate  final data values,  perform verification checks, and assign
flags to any detected aberrant values.

     The  program creates the Form 101 that summarizes  final data reported by the
preparation  laboratory.    Hard  copies  of  this  form are  sent  to all  DDRP
cooperating agencies and groups, in addition, the Form 101 is available on floppy
disk.
SAMPLE TRACKING

     The  preparation laboratory manager must track samples as they are received
from the sampling crews, during cold storage,  through processing and analysis,
and during batch  formation and shipment to the  analytical  laboratories.   The
computer data are used to track the location and progress of each sample while
it is physically at the laboratory.

     In  order to  track the sampling  activities  and shipment/receipt  of  bulk
samples, the following information is to be reported to the QA manager weekly:

         number of  samples received
     —   number of  pedons  sampled and percentage of total expected
     —   number of  sets received

     In order to track the progress of the preparation activities, the following
information is to be reported to the QA manager weekly:

         number and percent of samples that have completed specific analyses
         number of  bulk samples which are completely processed (excluding bulk
         density) and are  ready to be batched
         number of  sets that have been completed
         number of  batches sent to analytical laboratories


DATA VERIFICATION

     The  system  allows  the entry  and editing of data by preparation laboratory
personnel and the review of progress and data analysis by the QA staff.  After
a run of  data  is entered  into a  "temporary"  data file, the  final  values are
calculated and  verified.   A  printout of this  information  is checked  by the
preparation laboratory manager and by a representative from the QA staff.   The
data that have satisfied the verification criteria are then moved into a "master"
data file which is  used to track progress and percentages of the analyses  that
have been completed. The master file cannot be edited by preparation laboratory
personnel.

     The  five raw  data  forms that are entered into the computer are:  (1) sample
receipt raw data form,  (2) clod bulk density raw data form,  (3) known volume bulk
density raw data form,  (4) loss on ignition raw data form,  and (5)  bulk sample
processing raw data  form,   in the descriptions of the forms below, a list of the
data calculations,  verification  checks,  and pertinent verification  flags are
included.  At the beginning of  each description is a list of the data from each


                                 111

-------
particular form which must be transferred to  the  Form 101.   The flags consist
of two  characters,  an alpha character  and  a  numeric character.   The numeric
character designates which data  form the flag pertains to, e.g.,  a flag with the
numeric character "3" pertains to the known volume raw data form.


SUMMARY OF VERIFICATION FLAGS

Sample Receipt Raw Data Form

Al   Missing  data  on form
Bl   sample code mislabeled
Cl   site  ID  mislabeled
Dl   More  than 2 known volume samples
El   More  than 3 clods

clod Bulk Density Raw Data Form

A2   Missing  data  on form
B2   Submerged clod  weight >  Lab clod  weight,  and  clod did not float
C2   Clod  floated
D2   Problem  clod  (deletes value from  average)
E2   oven-dry clod weight > Lab  clod weight
F2   oven-dry clod weight > Field-moist  clod weight
G2   Coefficient of  variation >  20  among replicates
H2   Bulk  density  >  2.0
12   Bulk  density  <  1.0
J2   Sample code not recorded same  as  on sample  receipt form
K2   Number of replicates not recorded same  as on  sample receipt form
L2   Replicates >  3

Known Volume Bulk Density Raw Data Form

A3   Missing  data  on form
B3   Oven-dry weight > Air-dry weight
C3   Weight of fines + Rock fragments  >  Air-dry  weight
D3   Sample code not recorded same  as  on sample  receipt form
E3   Number of replicates not recorded same  as on  sample receipt form
F3   Bulk  density  >  2.0
G3   Bulk  density  <  1.0
H3   Coefficient of  variation >  20  among replicates

Loss on Ignition Raw Data Form

A4   Missing  data  on form
B4   Organic  matter  > 100%
C4   oven-dry weight > Air-dry weight
D4   Ashed weight  >  Oven-dry  weight
E4   sample code not recorded same  as  on sample  receipt form
F4   sample analysis duplicated
G4   organic  soil  type not correlating
H4   Mineral  soil  type not correlating
14   Internal duplicate/routine  pair > 15% RSD
J4   FAP samples > 15% RSD
K4   Field duplicate/routine  sample >  20% RSD
L4   FAL sample outside  accuracy window
M4   FAP sample outside  accuracy window
N4   Manager's sample outside accuracy window
                                 112

-------
Bulk sample Processing Raw Data Form

A5   Missing data  on form
B5   Mineral soil  air-dry moisture > 2.5%
C5   organic soil  air-dry moisture > 6.0%
D5   Field-moist pH > 6.0
E5   sample code not recorded same as on sample  receipt  form
F5   Internal duplicate/routine sample > 15%  RSD
G5   Field duplicate/routine  sample > 20%  RSD
H5   FAP samples > 15% RSD
15   pH Q.CCS buffer < 3.95 or > 4.05
J5   Manager's sample outside accuracy window
K5   FAP sample outside accuracy window


SAMPLE RECEIPT RAW DATA FORM

Form 101 Information;

SAM_CODE
SITE_ID
SET_ID

General;

     ERL-C provides EMSL-LV with the pedon  number (state,  county, unit) and site
ID for each  pedon  to be  sampled.   This information is placed in the data base
before preparation activities begin.

     All samples received from the field and their condition are recorded on the
sample receipt raw data form  and immediately entered into the data  base.  A file
comparison procedure is initiated, comparing the pedon number and site ID with
the raw  data from the sample  receipt form.   If  any  discrepancies  are found,
processing of the  affected samples ceases until the discrepancies are rectified.
All flags for the receipt data are unique and identified by the numeric character
" 1".
     After  all  verification  checks  for  the sample  receipt  data have  been
satisfied, the data are sorted by  set  ID.  All samples  with the same set ID will
undergo processing simultaneously.

Calculations;

     No calculations are to be performed on the  sample receipt  form.

Verification Flags

If any data are missing on the form; then flag Al
If sample code is  mislabeled; then flag Bl
If site ID is mislabeled; then flag Cl
If more than 2 known volume samples are recorded; then flag Dl
If more than 3 clods are recorded; then flag El

     When flags other than Dl and El are generated, processing of the affected
samples must cease until the problem is rectified.

Reporting;

     The  computer produces a printout, sorted by  SET_ID and SAM_CODE, that lists
all flags assigned.


                                 113 '

-------
CLOD BULK DENSITY RAW DATA FORM

Form 101 information;

BD_CLD

General;

     The clod data  are entered  into a temporary working file.  The verification
program should be run on a regular basis and the discrepancy printout generated.
All flags for the clod data are unique and identified by the numeric character
"2".

     After  all verification checks for the  clod  data have been satisfied,  the
data are sorted by  set ID and sample code.  All samples with the same set ID will
undergo processing simultaneously.

Calculations;

SARAN_WT (g) =  (FIELD_DP + LAB_DP) * (LAB_WT - FLD_WT) / LAB_DP

OD_SAR  (g)  = SARAN_WT * 0.85

SAJRAN VOL (cm3)  = SARAN_WT  /  1.30

R_FRAG VOL (cm3) =  R_FRAG  /  2.47

WATER VOL (cm3)  = CLOD_H20  /  WT_DEN

SOIL/PORES VOL  (cm3)  = WATER VOL - (R_FRAG VOL +  SARAN VOL)

BD_CLD (g/cm3) = [LAB_WT - (R_FRAG  + OD SARAN_WT + 1.88)] / SOIL/PORES VOL - 0.20

Verification Flags;

If any parameter field is blank, except for 'COMMENTS'; then flag A2
If CLOD_H20 > LAB_WT and FLOAT not marked 'Y';  then  flag B2
If FLOAT is marked "Y";  then flag C2
If PROS is marked  "Y";  then flag D2 (deletes value from average)
If FIELD_WT > LAB_WT; then flag E2
If CLOD_OD > FIELD_WT;  then flag F2
If CV > 20; then flag G2
If BD_CLD > 2.0; then flag H2
If BD_CLD < 1.0; then flag 12
If a SAM_CODE does not match the receipt file;  then  flag J2
If REP does not match the receipt file; then flag K2
If REP > 3; then flag L2

Reporting;

     For each SAM_CODE  there are usually  3 replicates,  although  this  number
varies.  Therefore, a check for replicate consistency is necessary.   If  the PROB
column is marked "Y" for any replicate  in  a  SAM_CODE,  this replicate should  not
be included in  the following calculation.
                                  114

-------
     The computer produces a printout, sorted by SET_ID,  SAM_CODE,  and REP, that
lists all flags assigned.  For each SAM_CODE, the mean,  standard deviation, and
coefficient of variation of all replicates are listed.

     The number of horizons and the  percentage  of  the total number of samples
having clods are reported.


KNOWN VOLUME BULK DENSITY RAW DATA FORM

Form 101 Information;

BD_KV

General;

     Known  volume samples are taken  from  those  horizons where clods cannot be
extracted.

Calculations;

FINES_WT (g)  =  (OD_WT - R_FRAG - BAG_WT)

BD_KV (g/cm3)   =  FINES_WT /  (VOL  - (R_FRAG / particle density)

verification Flags;

If any parameter field is blank, with the  exception of 'COMMENTS'; then flag A3
If OD_WT > AIR_WT; then flag B3
If FINES_WT + R_FRAG > AIR_WT; then flag C3
If a SAM_CODE does not match the receipt form;  then flag D3
If REP does not match the receipt form; then flag E3
If BD_KV > 2.0; then flag F3
If BD_KV < 1.0; then flag G3
If CV > 20; then flag H3

Reporting;

     The computer produces a printout, sorted by SET_ID,  SAM_CODE,  and REP, that
lists all flags assigned.  For each SAM_CODE, the mean,  standard deviation, and
coefficient of variation of all replicates are listed.

     The number of horizons and the  percentage  of  the total number of samples
having known volume samples are reported.


LOSS ON IGNITION RAW DATA FORM

Form 101 Information;

OM_LOI

Calculations;

     Each CRUC_NO will  be associated with  a pre-determined  CRUC_WT.

OM_LOI  =  [(OD_WT - ASHED_WT) / (OD_WT -  CRUC_WT)] * 100
                                 115

-------
verification Flags;

If any parameter field is blank,  with the exception of  'COMMENTS'; then flag A4
If OM_LOI > 100; then flag B4
If OD_WT > AIR_WT; then flag C4
If ASHED_WT > OD_WT; then flag D4
If a SAM__CODE does not match the receipt file; then flag E4
If the SAM_CODE is duplicated on the LOI form; then flag F4
If OM_LOI > 20 and SOILTYP on the bulk sample form is not "O"; then flag G4
If OM_LOI < 20 and SOILTYP on the bulk sample form is not "M"; then flag H4
If the internal duplicate/routine pair is > 15% RSD; then flag 14
If the FAP samples > 15% RSD; then flag J4
If the field duplicate/routine pair > 20% RSD; then flag K4
If a FAL sample is outside accuracy window; then flag L4
If a FAP sample is outside accuracy window; then flag M4
If a MS is outside accuracy window; then flag N4

Reporting;

     The computer  produces a printout, sorted by SET_ID  and SAM_CODE, that lists
all flags assigned.

     The  number of horizons and  the  percentage  of  the  total number of samples
that have been analyzed are reported.


BULK SAMPLE PROCESSING RAW DATA  FORM

Form 101 Information;

PH_MP
MOIST_P (1,2,3)
RF_FG
RF_MG

General;

     General information concerning the bulk  sample  is recorded on this form.
when the  form is complete, the  sample  is  considered  ready  to  be batched and
shipped to an analytical laboratory.

Calculations;

PH_MP  (no calculation required)

MOIST_P (1, 2, 3)  =  [(INIT_WT1 - OD_WT /  (OD_WT - TIN_WT)]  *  100

     Each  MOIST_P calculation will  be conducted  only  if the  previous MOIST_P is
> 2.5 percent for mineral samples or > 6.0 percent for  organic samples.

RF_FG  =   (RF_FGRAW / A_BULK)  * 100

RF MG  =   (RF MGRAW / A BULK)  * 100
                                   116

-------
Verification Flags;

If any parameter field is blank,  with the exception of  'COMMENTS'; then flag A5
If SOILTYP = 'M' and MOIST_P (1,2,3) > 2.5; then flag B5
If SOILTYP = 'O' and MOIST_P (1,2,3) > 6.0; then flag C5
If PH_MP > 6.0; then flag D5
If a SAM_CODE does not match the receipt file; then flag E5
If the internal duplicate/routine pair > 15% RSD; then flag F5
If the field duplicate/routine pair  for MOIST_P, RF_FG,  or RF_MG >  20% RSD; then
   flag G5  (or >0.20 SD for PH_MP)
If the FAP samples for MOIST_P,  RF_FG, or RF_MG > 15% RSD; then flag H5  (or
   0.15 SD for PH_MP)
If pH QCCS out of window; then flag 15
If MOIST_P is outside accuracy window; then flag J5
If PH_MP is outside accuracy window; then flag K5

Reporting;

     The computer produces a printout, sorted by SET_ID  and SAM_CODE, that lists
all flags assigned.

     The  number of horizons and the percentage  of the  total  number of samples
that have been analyzed are reported.
                                 117

-------
                              Appendix D

  A Comparison of Soil Sample Homogenization Techniques
      The following study was initiated to determine if the homogenization technique used at the
different Direct/Delayed Response  Project  preparation laboratories was a  major source of
measurement uncertainty.
                                   118

-------
                              Internal Report


A COMPARISON OF SOIL SAMPLE HOMOGENIZATION TECHNIQUES
                                   by
              B. A. Schumacher, K. C. Shines, J. V. Burton, and M. L Papp
                 Lockheed Engineering and Sciences Company, Inc.
                          Las Vegas, Nevada 89119
                          Contract Number 68-03-3249
                               Project Officer

                                L J. Blume
                     Exposure Assessment Research Division
                   Environmental Monitoring Systems Laboratory
                        Las Vegas, Nevada 89193-3478
          ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
                OFFICE OF RESEARCH AND DEVELOPMENT
               U.S. ENVIRONMENTAL PROTECTION AGENCY
                     LAS VEGAS, NEVADA  89193-3478
                                119

-------
                                       Notice


       This document is  intended for internal Agency use only.  Mention of trade names or
commercial products does not constitute endoresement or recommendation for use.
                                      120

-------
                              ACKNOWLEDGEMENTS
       We greatly appreciate the assistance of Gregory A. Raab, William H. Cole III, Conrad A.
Kuharic, Gerald L Byers, Rick D. Van Remortel, and Robert L Tidwell in the technical review of this
document.  Our thanks are also extended to Mohammad J. Mian  for his aid in producing and
interpreting the statistical analyses.
                                      121

-------
                                     Introduction
       The need for sample homogeneity prior to laboratory analyses has been long recognized by
geologists, pedologists, chemists, and members of other scientific disciplines. Homogeneity is the
degree that the material under investigation  is mixed resulting in the random distribution of all
particles in the sample.  Completely homogenous materials are so rare that they may be considered
nonexistent (Ingamells and Pitard, 1986) yet scientists must strive to obtain a homogenous sample
in order to obtain data exhibiting minimal error attributable to sample heterogeneity.

       Care must be taken during the subsampling phase of soil preparation, during sample
transport, unpacking, and transfer to other containers in the laboratory to avoid particle sorting via
different particle densities, shapes, sizes, and resistance of certain minerals to mixing, such as
magnetite (Muller, 1967; Ingamells and Pitard,  1986; Reeves and Brooks, 1978). Numerous methods
have been  used to obtain homogeneous soil samples. The  various methods range  from simple
grinding and sieving of the sample to a desired particle-size to various mixing and splitting devices
and  machines.   Each method  will  be  discussed  individually  with both  advantages   and
disadvantages being presented.

       One of the most common methods used to obtain a homogeneous sample is  to grind  and
sieve the soil to a desired  particle-size (generally <2-mm) followed by random sampling.  This
method assumes that the initial material is ground without preference to any given factor, such as
color, and  that during grinding and sieving, the sample becomes sufficiently homogenized.  To
further eliminate possible heterogeneity within the sample and to reduce the sample size to the
desired quantity for a given analysis, subsamples may be obtained by "spooning" (Carver, 1981)
or some other method which involves the random insertion of a spoon or other sampling device  into
the previously ground and sieved sample.  This method is preferably done rapidly  and without
extensive visual examination of the sample that could lead to  a processor preference in certain
cases, for example, light catching the shiny surfaces of mica flakes leading to preferential inclusion
or exclusion of that part of the sample. The "spooning" type of subsampling markedly reduces the
time of sample processing in comparison to multiple, successive splitting operations and has been
shown to produce equivalent correlation coefficients between observed settling velocities of sands
(median r-0.994 using other splitting methods and median r=0.9955 using the spooning technique)
when used to obtain 20 gram subsamples from initial 1 to 2  kg sand  samples (Carver, 1981).

       Two other methods have been devised which are similar to the "spooning" method except
they involve an intermediate step between grinding and  sieving and  subsampling  procedures.
Gilliam and Richter (1985) used an intermediate step of stirring the sample with a spatula before
subsampling occurred, presumably until visual homogeneity was obtained. Some analysts content
themselves by merely shaking the sample in  a bottle prior to subsampling  and ignore the risk of
constituent segregation (Reeves and Brooks,  1978).

       An  additional method of sample  subdivision  with the goal of  obtaining a representative
sample was presented  by Allman and Lawrence (1972). Their method is similar to the "spooning*
method of  Carver (1981) except that a scoop  of ground and sieved materials was divided among
four containers.  The process was repeated continually changing the filling order of the containers
until the sample had been quartered or an appropriate sample size had been obtained. As with the
spooning process, visual bias as to how much sample and into  which container the samples were
placed is a concern.

       Before proceeding to the more elaborate sample splitting  schemes, a discussion of methods
used simply to mix (homogenize) the ground and sieved sample is warranted. The simplest of the
homogenization processes  is tumbling the sample on a sheet of paper, cloth, or plastic.  This
process involves the manual rolling of ground samples such that the  sample must tumble upon
itself and  not just slide along the surface of  the sheet (Schuler, 1971; Van Johnson and Maxwell,
1981; Ingamells and  Pitard,  1986).  This method is effective on sample sizes less than 2 kg, yet
caution must be taken to ensure that the sheet material does not contain any element which is to


                                       122

-------
                                Fig. 1. A twin shell V-blender.
be quantitatively analyzed and does not develop static charges that may lead to segregation of the
finer particle sizes.

       Homogenization may also be achieved through  the use of mechanical mixing devices
including a spiral mixer, a cement mixer, and a twin shell V-blender.  The spiral mixer involves the
rotation of the bottled sample in both horizontal and vertical planes (Van Johnson and Maxwell,
1981).  The cement mixer or similar devices involve  the rotation of the sample  in a chamber with
a series of internal baffles  that cause the materials to be thoroughly tumbled and mixed.  These
two methods are useful for samples ranging from less than one pound to several hundred pounds.
The twin shell V-blender involves the rotation of two hollow cylinders about an horizontal axis such
that  the  apex describes a circle in the vertical plane  (Schuler,  1971).   Twin shell  blenders are
available commercially in sizes ranging from 4 to 16 quarts (approximately 10 to 40 kg of a mineral
soil)  internal capacity (Fig. 1). However, Ingamells and Pitard (1986) expressed serious doubts as
to the value of mechanical splitters. They stated that, "In general, machines that use mechanical
violence and look and sound as though  they are efficient are most likely to cause segregation of
heavy and light,  large and  small, and  flat and round particles*.  Further concerns have  been
expressed concerning sample breakdown to finer particle sizes due to the violent tumbling in the
machines.

       The  riffle  splitter (also called  a  chute  splitter, Jones splitter, or just  sample splitter) is
perhaps  the most common mechanical  method  for sample homogenization and/or  sample
sizereduction (Ingram, 1971, Mullins and Hutchison, 1982).  The riffle  splitter also provides one of
the best general  methods of sample mixing to obtain bulk sample homogeneity (Ingamells and
Pitard,  1986).  A riffle splitter is a device having an equal number of narrow sloping chutes with
alternate chutes discharging the sample in opposite directions into two collection bins (Figs. 2 and
3).  Sample homogenization is  achieved  by  repeated  pouring of soil  through the splitter  and
combining the halves between passes.  The use of the riffle splitter as a  subsampling device is
done in a similar  manner with the exception that after the sample is passed through the splitter,

                                         123

-------
                                                 60° slope
                                                 on chutes/1
                          knife  edges


                          alternate openings

                        /collecting bin
                                                 collecting
                                                    bin
         Fig. 2. An open-bin riffle splitter.
Fig. 3. A closed-bin riffle splitter.
one collection pan is replaced with a clean pan. The material in the "replaced" pan, which contains
about one-half of the original sample, is then passed through the riffle splitter again thereby
reducing the volume in the clean pan to one-quarter of its original sample volume.  This process
of sample reduction is repeated until the desired weight or sample size is obtained.

       Many variations on the size and construction materials have been built on the principle of
the riffle splitter.  For small samples (10 to 50 grams), an efficient microsplitter has been designed
by Newton and Dutcher (1970) using balsa wood and glass microscope slides.  These authors
conducted experiments using 40 gram samples of the fine sand fraction and found only an 1.2%
average error when quartering the sample by a 2-step halving.  The results of Newton and Dutcher
(1970) are supported by the earlier work of Griffiths (1953) who found that sample homogenization
using a riffle splitter produced an  overall coefficient of variance  (Cy)  between median grain sizes
of less than 3% in 9 out  of 10 samples run on various rock and sand samples.  In the one case
where the Cv was greater than 3%, the error was attributed to operator differences (16 in the single
case vs. 5 or less in the other nine cases). A portable sieving and splitting device for field use has
been designed by Ibbeken (1974) which uses the riffle splitter to subsample and process coarse
grained sediments.  During his  research on unconsolidated sediments, no significant differences
were found between splits in terms of petrographic mineralogical composition, with the exception
of the  125  to 160 mm serpentine  fraction  (represented by only four pebbles), nor in grain size
distribution.  Another variation of the riffle splitter found in the literature was the use of a single
piece  of  tin-plate bent several  times to form a riffle splitter (McKinney and Silver,  1956).  The
advantage of this construction method is that free grain flow in the chutes was obtained, without
the hindrance of unevenly soldered joints that may be present in other riffle splitters.

       The use of riffle splitters and their variations are valuable in splitting samples  which range
in size from less than 10 grams  (Humphries, 1961) to several kilograms (Van Johnson and Maxwell,
1981). Ibbeken (1974) reported that 0.5 to 1 ton (455 to 909 kg) of sample can be processed daily
using a standard riffle splitter to reduce the initial quantity to 5 kg subsamples.  Although most
authors  find the  use of  the  riffle splitter to be  an effective,  efficient  method   for sample
homogenization and sample splitting, several problems exist.

                                         124

-------
       The major source of error involved in using a riffle splitter is the loss of fine particle sizes
via "dusting" when processing air-dried or oven-dried samples which contain fine particle sizes (Van
Johnson and Maxwell, 1981).  The  process of riffle  splitting requires that a uniform stream of
material be poured  into the mouth or top of the splitter.  Dust loss may occur through the chute
ends, in the collection bins, and through the mouth openings (Fig. 2) in an open-bin system or just
through the mouth if a closed-bin system  is used (Fig. 3).

       Several  studies have been conducted comparing the effectiveness of the riffle splitter to
other homogenization techniques. Wentworth et al. (1934) compared the riffle splitter with a rotary
splitter (to be discussed) and found that the rotary splitter more accurately split an initial sample
of known grain size distribution into subsamples with similar grain size distributions than the riffle
splitter. Kellagner and Flanagan (1956) in a comparison experiment among a multiple-cone splitter
(to be discussed), cone and quartering (to be discussed), and a riffle-based microsplitter found the
microsplitter was the worst for both accuracy and precision of grain frequency percentages for
subsampling three different weights (5, 10, and 20 grams) of an artificially created very coarse and
coarse sand fraction mixture.  In a  similar type of study, Mullins and Hutchison (1982) compared
the Cv among sand fraction contents of several soils and ranked, in order of best to worst, the
rotary subsampling, riffle splitting, cone and quartering, and spoon sampling, in terms of their ability
to homogenize a sample. These authors did note, however, that the best and worst methods were
only significantly different at the 10  percent level.

       Perhaps the best  known sample  splitting  method is the classical cone  and quarter
technique.  This technique involves pouring the sample into a cone, flattening the cone, dividing the
flattened cone into  four equal divisions (quartering), and then removing 2 opposite quarters (Fig.
4). The remaining two quarters are replied into a cone and the process is repeated until the desired
sample size is obtained.  Variations on the process are possible which can enhance the  speed of
sample  size reduction by using just one quarter (chosen  at  random) to continue the splitting
process or which allow this method to be modified to homogenize a large sample.  The use of the
cone and  quarter method to homogenize a sample involves the removal of the  first quarter and
repiling it  into a cone followed by the  subsequent repiling of the opposite quarter and then the
remaining two  quarters to  reform a single cone (Raab et  al., 1990).  This process  is  repeated
several times until sample homogeneity is achieved.

       Several  sources of error  for  this method have  been identified.  Van Johnson and Maxwell
(1981) reported  that during  the cone and quarter process on large samples (several kilograms),
there is the danger of unequal  segregation of heavier materials during the flattening and coning
of the sample.  Similar to the riffle splitting techniques, dusting is also a possible source of error
during cone formation. Sample loss from the inability to recollect all the soil from the underlying
material, the  ability of the  sample  to  "cling" to the underlying material via static charges, and
sample embedment are all  further sources of error.  Muller (1967)  placed a limitation on  the cone
and  quarter technique to samples greater than 50 grams and stated that this method is most
successful in the  field for sample processing of larger sample sizes. Raab et al., (1990) used the
cone and quarter method to homogenize large volumes of soil («150 kg) and then subsampled the
repiled cone with a 2-foot long plastic tube (2-inch internal diameter). They then sieved the sample
into  3 particle size classes  (2-mm to 0.105-mm, 0.105-mm to 0.053-mm, less than 0.053-mm) and
subsampled using a riffle splitter. A resieving to check particle class weights of the three divisions
between the six subsamples resulted in the finding that no significant difference (error level = .05)
among the subsamples indicating a homogenous mixture had been obtained.
                                            125

-------
                           Fig. 4. The cone and quarter technique.
       Several other sample splitting/homogenizing devices found in the literature include a rotary
splitter (Muller, 1967; Schuler, 1971; Allman and Lawrence, 1972), a brass disc microsplitter (Brewer
and Barrow, 1972), and a multiple cone splitter (Kellagher and Flanagan, 1956).  The rotary splitter
involves the pouring of the sample into a feeding hopper located above a rotating disc containing
sample bottles or pans (Fig. 5). The disc is mechanically rotated and the sample is divided among
the collecting bottles or pans. The brass disc microsplitter was designed by Brewer and  Barrow
(1972) for subsampling  small paniculate samples which range  up  to  1 gram in the particle size
range between 10 and 200 microns.  This splitter divides the sample by passing it through a fluted
brass disc which contains several holes to separate the sample into  different collection bottles.
The device  also has a mechanical vibrator  system to ensure complete  collection  of the fine
particles.  The multiple-cone splitter of  Kellagher and Flanagan (1956) consists of three funnels
and brass cones vertically mounted which are capable of splitting small samples (10 to 400 grams)

into 5 mg subsamples which are collected in one of four divisions in the base pan (Fig. 6).  These
authors statistically compared the precision and accuracy of grain frequency percentages obtained
using their device with a riffle-based microsplitter and the  cone  and quarter technique and found
that their splitter was better in both categories than the riffle splitter while being better than the
cone and quartering in terms of accuracy only.

       Due to the overwhelming concurrence  (although not unanimous) in the literature as to the
value  of the cone and quarter and  riffle splitting techniques in sample homogenization  and the
common use  of random sampling  after grinding  and sieving  of  the soil,  investigations  were
undertaken  to  determine   the  effectiveness  (degree of  homogenization),  efficiency  (time
consumption), and the  extent  of  the loss  of fine particles  by these methods  during the
homogenization of large soil samples.
                                         126

-------
    Fig. 5. A rotary splitter.
             i      /                >^
             »  I   / >M»BIL |rr»li>f»
Fig. 6. A multiple-cone splitter.

-------
                                Materials and Methods

       Bulk samples (approximately  15 kg) of  surface  horizons  of the Overton  clay  (fine,
montmorillonitic, calcareous, thermic Mollic  Haplaquepts), Gila  silt  loam (coarse-loamy, mixed,
calcareous, thermic Typic Torrif luvents), Calico loam (coarse-loamy, mixed calcareous, thermic Typic
Torrifluvents), and Jean gravelly loamy fine sand (sandy-skeletal, mixed, thermic Typic Torriorthents)
were collected in Clark  County, Nevada (Speck, 1985).  Surface horizons with different textures
were selected to represent soil systems with varying sensitivities to the  loss of fines that may
occur during sample homogenization.

       Moist bulk samples were split into three subsamples  (approximately 5 kg each) and air-
dried.  Samples were ground  and sieved through a 2-mm (10  mesh)  sieve with three random
samples being collected from each of the subsamples  prior to  any  homogenization procedures.
The three original subsamplos from each soil series were "homogenized" by either passing the soil
seven times through the cone and quarter technique, an open bin  riffle splitter, or a closed bin riffle
splitter.  Three samples of approximately 60 g each were collected after the first, third, and fifth
passes.  Seven samples were collected after the seventh pass through each of the  homogenization
processes. Samples were collected from the riffle splitters by placement of the receiving bottle
under randomly selected chutes  prior to the sample being poured through the splitter.  Samples
from the cone and  quarter technique were collected by pouring the sample  over the collection
bottles during cone  rebuilding and subsequently removing  the bottles when they were full prior to
completion of the cone using all  four quarters.

       A timed experiment was conducted to determine the efficiency of each splitting technique
and  involved the passing of the previously "homogenized" soil (from the seventh pass) through the
three homogenization  procedures seven times without  collection  of soil samples after  any
intermediate passes.  Seven samples were collected after the timed experiment in which the soil
had  now been passed fourteen times through the homogenization process.

       Three parameters  were selected to  determine the effectiveness  of the  homogenization
method and loss of fines from the system, namely, particle-size analyses (for loss of inorganic
fines), loss on ignition (LOI) organic matter (for organic  fine losses)  and pH (for bulk chemical
changes).  Particle-size distribution (<2-mm) was determined usjng the pipette method described
by Gee and Bauder (1986). Soil samples were oven-dried  at 110° C overnight and loss on ignition
organic matter content was determined gravimetrically after heating at 450  C for a minimum of
6 hours. The pH values were determined in a 1:2 soil:0.01 MCaC\2 solution ratio (McClean,  1982).
Levels of confidence were determined by analysis of variance (ANOVA).

                                Results and Discussion


        It should be noted that pass no. 0 represents the random samples  collected prior to any
sample homogenization process other than preparatory grinding and sieving.  Complete data tables
containing individual analyses of sand, silt, clay,  sand fractions, pH (no standard deviations or
%RSDs presented),  air-dry moisture contents, and LOI organic matter as well as means , standard
deviations, and %RSDs for replicate and cumultive (for all samples in the given treatemtn of a given
texture) data are presented in the Appendix

Observed Sources of Soil  Loss


        Loss  of fines via dusting was most apparent during the  use  of the open bin riffle splitter.
Dust loss occurred from the mouth of the riffle  splitter due to the air-dried soil hitting the baffles
and sliding down the  chutes  as well  as in the collecting bin  from the soil falling  upon itself.
Apparent fine particle losses from use of the closed bin riffle splitter  were less noticeable than for
the open bin riffle splitter.  Dust losses occurred through cracks within the riffle  splitter and through
the  mouth into which the soil was being poured. However, although the overall dust loss appeared
to be less, additional soil loss was observed within the closed bin riffle splitter  where soil collected

                                         128

-------
on internal ledges and around the outside of the collecting bins. Little loss of fines, via dusting,
was observed using the cone and quarter technique. Dusting occurred only during the piling of the
quarters upon each other to form a new cone. Soil losses were observed due to an inability to
completely transfer all soil materials during new cone formation.

Influence of Sample Homogenization on pH

       No  significant differences were found in pH values among replicates  regardless of soil
texture, splitting method, or the number of the pass from which the subsample was obtained (Table
1). All replicate pH values met the intralaboratory precision goals set for the Mid-Appalachian soi
survey (0.10 pH unit).   The ranges in pH for the Jean gravelly loamy fine sand and Calico sandy
loam were  7.7 to 7.8.  The Gila silt loam had  pH values of 7.8 in all samples while the Overton clay
had  pH values ranging from  7.9 to 8.0.

Influence of Sample Homogenization on Particle-Size Distribution

       All replicate total sand, silt, and clay contents met the intralaboratory precision goals set
for  the  Mid-Appalachian soil survey (3.0  wt% standard  deviations for sand  and silt; 2.0  wt%
standard deviations for clay).  Standard deviations for total sand,  silt,  and clay contents ranged
from 0.05 to 1.30, 0.05 to 2.59, and 0.00 to 2.00, respectively (Table  2).

       The lowest standard deviations among replicates were almost always found after the
samples had been  passed five times  through  the homogenization process  regardless of
homogenization technique used or soil texture (Table 2). When exceptions did occur, the standard
deviations were not significantly different  from the fifth pass and  occurred in  either  the random
sample or after the first pass through the homogenization process in which the sample was simply
halved via riffle splitting or was effectively a random sample from the first flattened quarter for the
cone and quarter  technique.  At the 90% or greater  confidence interval, the random (pass 0), first,
third, seventh, and timed experiment passes had 36, 33, 53, 86, and 97% of the samples having
significantly greater replicate variances than the fifth pass, respectively.  If  a 75% or  greater
confidence limit was used, pass  0,  1, 3,  7, and T had 53, 56, 78, 98, and 100% of the samples,
respectively, had significantly greater variances than the fifth pass.  Earlier passes  (pass 0, 1, and
3) resulted in less samples with significantly different variances  among  the  replicates  than in
passes after the fifth, yet one-third to  more than one-half of the samples had significantly greater
variances than found after five homogenization passes. These data indicate that any attempts to
further homogenize the soil after the fifth pass through the various homogenization methods
                                       129

-------
Table 1. Soil pH values determined in a 1:2 soil:solution
        of 0.01 M CaCI. for homooenization studv*.
Split No.

0
1
3
5
7
T

0
1
3
5
7
T

0
1
3
5
7
T

0
1
3
5
7
T
Open Bin Closed Bin Cone and Quarter
Riffle Solitter Riffle Solitter Techniaue

7.7
7.7
7.7
7.7
7.7
7.7

7.7
7.7
7.7
7.8
7.7
7.7

7.8
7.8
7.8
7.8
7.8
7.8

7.9
7.9
7.9
7.9
7.9
7.9
Jean loamy fine sand
7.8
7.7
7.7
7.7
7.7
7.7
Calico sandy loam
7.8
7.7
7.8
7.7
7.7
7.8
Gila silt loam
7.8
7.8
7.8
7.8
7.8
7.8
Overton clay
8.0
8.0
8.0
8.0
8.0
8.0

7.8
7.7
7.7
7.7
7.7
7.7

7.7
7.7
7.7
7.7
7.7
7.7

7.8
7.8
7.8
7.8
7.8
7.8

7.9
7.9
7.9
7.9
7.9
7.9
a = data presented are the means of 3 replicates for splits 1
    through 5 and means of 7 replicates for splits 7 and T.
                           130

-------
 Table 2. Standard deviations between replicate analyses for particle-size distribution and loss on
 	ignition organic matter*.	
  0
  1
  3
  5
  7
  T
              Open Bin
             Riffle Splitter
 Closed Bin
Riffle Splitter
Cone and Quarter
   Techniaue
Split     Sand"   Silt   Clav   O.M.   Sand   Silt   Clav   O.M.   Sand   Silt   Clav   P.M.
                                      Jean loamy fine sand
                                                                                          -)
0.15*"
0.12"
0.22"*
0.05
0.18"*
0.17*"
0.08
0.20*"
0.24*"
0.06
0.13"*
0.21*"
0.15*
0.08
0.08
0.08
0.16"*
0.1f"
0.05*"
0.01
0.03***
0.01
0.02"*
0.03*"
' 0.09
0.19*
' 0.73"*
0.09
' 0.25*"
' 0.22"*
0.05
0.16
0.63***
0.11
0.23**"
0.24*"
0.08
0.11
' 0.11
0.09
' 0.10*
0.13*"
0.07*
0.03
0.07*
0.04
0.14"*
0.16*"
0.33*"
0.10
0.19*
0.10
0.18"*
0.30"*
0.38*"
0.06
0.25"
0.10
0.15*"
0.35*"
0.08'"
0.04*
0.07"
0.02
0.07*"
0.10*"
0.05"
0.04*
0.03
0.02
0.03"
0.01
  0
  1
  3
  5
  7
  T
0.10
0.64"
0.89*"
0.23
0.30"
0.46"*
0.36"
0.63*"
0.79*"
0.15
0.32*"
0.74*"
0.41"
' 0.22
' 0.20
0.16
' 0.30*"
' 0.52*"
0.10*"
0.03
0.06"
0.02
0.04*"
0.10*"
0.34
0.52*
0.40
0.28
0.34"
0.33"
0.34
0.43
0.60*
0.33
0.36*
0.45"
0.00
0.10
0.21*
0.09
0.42*"
0.35*"
0.07*
0.07*
0.06*
0.03
0.11"*
0.06*"
0.27*
0.29"
0.16
0.12
0.47"*
0.54"*
0.48*"
0.13*
0.25*"
0.07
0.34*"
0.52"*
0.31*"
0.16*
0.11
0.08
0.34*"
0.24*"
0.08'
0.05'
0.02'
0.01
0.05'
0.07'
  0
  1
  3
  5
  7
  T
0.31*
0.20
0.31*
0.18
0.34*"
0.83"*
0.41*
0.19
0.26
0.19
0.51"*
0.88"*
0.21
0.19
0.39"
0.17
0.28*"
0.43*"
0.02*
0.04*"
0.06*"
0.01
0.07*"
0.08*"
0.37
0.48*
1.30*"
0.29
0.56*"
0.37"
0.30
0.17
0.96*"
0.25
0.49**"
0.42*"
0.26
0.32*
' 0.35*
0.21
' 0.22*
' 0.23*
0.09***
0.08"*
0.06"
0.02
0.06"*
0.07*"
0.67*"
0.23
0.50"
0.19
0.19
0.40***
0.38*"
0.06
0.51*"
0.11
0.34*"
0.28"*
0.46'"
0.28**
0.14
0.12
0.23*"
0.39*"
0.03
0.04
0.04
0.03
0.07'
0.07'
                                         Overton clay
0
1
3
5
7
T
1.11"*
0.83"
0.61*
0.28
0.71"*
0.53*"
0.57
2.59*"
2.07*"
0.58
1.02"*
233*"
0.56*
1.85*"
1.46*"
0.31
0.57*"
200*"
0.05
0.07*
0.09*
0.04
0.07*"
0.05"
0.32
0.07
0.51*
0.29
0.32*
0.53*"
0.24*
0.21*
0.39"
0.13
0.37"*
0.65*"
0.30
0.26
0.90"*
0.24
0.47"*
0.59*"
0.24***
0.09*"
0.12*"
0.02
0.06"*
0.08*"
0.25" 0.17
0.59*" 0.66"
0.20" 0.50*
0.09 0.23
0.51"* 0.38*"
0.42"* 0.85"*
0.08
0.55"
0.44*
0.20
0.46*"
0.66*"
0.10""
0.06*
0.08"
0.03
0.07"*
0.10*"
a - data presented are the standard deviations of 3 replicates for splits 1 through 5 and means
    of 7 replicates for splits 7 and T.
b - * = 75%; ** = 90%; *** = 95% confidence levels as determined by ANOVA.  Significant
    differences are compared to the fifth pass.
                                          131

-------
markedly increased the variability among replicates and  thus the  heterogeneity of the sample
regardless of soil texture or homogenization technique.

Influence of Sample Homoqenization on Loss on Ignition Organic Matter

       The lowest standard deviations among replicates for LOI organic matter content were found
after the fifth pass,  with only two exceptions, similar to the results for  standard deviations in
particle-size distribution (Table 2).  The two exceptions were noted in the Jean gravelly loamy fine
sand  (after  the  first pass through the closed-bin riffle splitter and after the  timed experiment
homogenized by the cone and quarter technique), yet neither standard deviation was significantly
less than the standard deviation of the fifth pass. When  a confidence interval  of 90% or greater
was established, 58, 33, 50, 100,  and 91% of the replicates from passes 0, 1, 3, 7, and T,
respectively, had significantly greater variances than those found in  the fifth  pass.  At the 75% or
greater confidence level, two-thirds or more of the replicates (83, 66, 83, 100, and 91% for passes
0, 1, 3, 7, and T, respectively) had significantly greater variances. These results support our earlier
findings that the sample appears to have achieved the greatest homogeneity after the fifth pass
through the  homogenization process.

Homogenization  Technique Efficiency

       The average time required to homogenize a sample using seven passes was 5.00 (range
4 to 5 minutes),  5.50 (range 4 to 7 minutes), and 30.75 minutes (range 24 to 44 minutes) for the
closed-bin riffle splitter, open-bin riffle splitting, and cone and quartering, respectively (Table 3). The
slightly longer time required to perform open-bin riffle splitting compared to closed-bin riffle splitting
was attributed to an initial lack of experience in the use of a riffle splitter by the technician.
Excluding the first time use of either riffle splitter, the average time required  for open-bin  riffle
splitting decreased to 5.00 minutes per sample compared to 4.3 minutes per sample required for
the closed-bin riffle splitter.  Excluding the first  use of the cone and quarter  technique  by the
technician, the average time required to  homogenize  the  sample decreased from  30.75 to 26.33
minutes per sample.  These results indicate that the use  of the riffle splitter for homogenization
required markedly less time per sample than samples  homogenized  by the cone and quarter
technique and was thus more efficient.
                                          132

-------
Table 3. Efficiency of various homogenization techniques.

        Open Bin    Closed Bin   Cone and Quarter
Soil    Riffle Splitter  Riffle Splitter      Technique
       (	minutes	)
Jean
Calico
Gila
Overton
Mean
Mean-1b
4.
7.*
6.
5.
5.50
5.00
4.
4.
5.
5.*
4.50
4.33
24.
44.*
29.
26.
30.75
26.33
a - * = first sample homogenized by the given method.
b - Mean-1 = mean of timed experiment excluding first
    sample homogenized by the given method.
Table 4. Loss of fines between the seventh pass and
        timed experiment.
Open Bin
Soil Riffle Solitter
Jean
Calico
Gila
Overton

-0.03'
-0.02
+0.09
-0.36
Closed Bin Cone and Quarter
Riffle Splitter Techniaue
%fslf)\/« .
nc
-0.10
-0.22
-0.19
\
+0.06
-0.56
-0.57
-0.42
a - +  = clay increase; - = clay loss; nc = no change.
                           133

-------
Loss of Fines

       Loss of fines was determined by comparison of clay contents between the seventh split and
timed experiment. Clay contents generally decreased after the additional seven passes through the
homogenization procedures but the overall clay content loss was very small (less than 0.6%) and
could be attributed to expected variance in the homogenization and analytical methods (Table 4).
It was interesting, however, to note that the greatest clay content decreases were found when the
soils underwent cone and quartering as the homogenization process. This result was due to a two-
fold effect in which the inability to recover all the soil from the underlying paper led to a greater
clay  loss from gap filling  between the larger sand particles and  perhaps due  to static charge
development on the paper leading to a retention of the charged clay particles.

                                      Conclusions

       The use of riffle splitting to homogenize a bulk soil sample was more efficient and had less
loss  of  fines  than cone  and quartering  and is  therefore  the recommended  homogenization
technique.  The use of a closed-bin riffle splitter was preferred to an open-bin riffle splitter due to
its greater apparent ability to contain and reduce the loss of fines from dusting.  Only five passes,
instead of seven, should be used to obtain the most  homogeneous sample, in terms  of particle-
size distribution  and  loss on  ignition organic matter,  due to  the overwhelming evidence that the
least variability among replicates occurred after the fifth  pass using all three splitting  techniques
and all four soil textures.   Random sampling after grinding  and sieving was the most efficient
homogenization method, since no additional sample preparation was involved, yet these samples
almost consistently had greater  replicate variabilities than the other  homogenization  techniques
after the fifth pass reducing the value of this technique for soil sample homogenization.
                                        134

-------
                                       REFERENCES
Allman, M. and Lawrence, D.F. (1972). Geological Laboratory Techniques. ARCO Publishing Co., Inc.
 New York. 335 pp.
Brewer, R. and Barrow, K.J. (1972). A Microsplitter for Subsampling Small Paniculate Samples. J.
 Sed. Pet. 42:485-487.
Carver, R.E. (1981). Reducing Sand Sample Volumes by Spooning. J. Sed. Pet. 51:658.
Gee, G.W. and Bauder, J.W. (1986). Particle Size Analysis. In A. Klute (ed.) Methods of Soil Analysis,
 Part 1. Agronomy 9:383-411.
Gilliam, F.S. and Richter, D.D. (1985). Increases in Extractabte Ions in Infertile Aquults Caused by
 Sample  Preparation. Soil Sci. Soc. Am. J. 49:1576-1578.
Griffiths, J.C. (1953). Estimation of Error in Grain Size Analysis. J. Sed. Pet. 23:75-84.
Humphries, D.W.  (1961). A Non-Laminated Miniature Sample Splitter. J. Sed. Pet. 31:471-473.
Ibbeken, H.  (1974). A Simple Sieving and Splitting Device for Field Analysis of Coarse Grained
 Sediments. J. Sed. Pet. 44:939-946.
Ingram, R.L (1971). Sieve Analyses. In. R.E. Carver (ed.) Procedures in Sedimentary Petrology. J.
 Wiley and Sons, Inc. New York. pg. 49-67.
Ingamells, C.O. and Pitard, F.F. (1986). Applied Geochemical Analyses. J. Wiley and Sons, Inc. New
 York. 733 pp.
Kellagher. R.C. and Flanagan, F.J. (1956). The Multiple-Cone Splitter. J.  Sed. Pet. 26:213-221.
McClean, E.O. (1982). Soil pH and Lime Requirement. In A.L Page (ed.)  Methods of Soil Analysis,
 Part 2. Agronomy 9:199.223.
McKinney, C.R. and Silver,  LT. (1956). A Joint-Free Sample Splitter. Amer. Mineralogist 41:521-523.
Muller, G. (1967).  Methods in Sedimentary Petrology. In W. Van Engelhart et a!., Sedimentary
 Petrology, Part 1. Hafner Publishing Co., New York. 283  pp.
Mullins, C.E. and  Hutchison,  B.T. (1982). The  Variability Introduced by Various Subsampling
 Techniques. J. Soil Sci. 33:547-561.
Newton,  G.B. and Dutcher, R.R.  (1970). An  Inexpensive Student Sample Splitter. J. Sed. Pet.
 40:1051-1052.
Raab, G.A., Bartling, M.H. Stapanian, MA,  Cole, W.H.,  Tidwell, R.L and  Cappo, K.A. (1990). The
 Homogenization  of Environmental Soil Samples in  Bulk.  In M. Simmons (ed.) Hazardous Waste
 Measurements. Lewis Pub. In press.
Reeves, R.D. and  Brooks, R.R. (1978). Trace Element Analysis of Geological Materials. J. Wiley and
 Sons, Inc. New York. 421  pp.
Schuler, V.C.O. (1971). Chemical Analysis and Sample Preparation. In R.E. Wainerdi and E.A. Uken
 Ed.  Modern Methods of Geochemical Analysis.  Plenum Press, New York. 397 pp.
Speck, R.L (1985). Soil Survey of Las Vegas Valley Area,  Nevada. USDA-Soil Conservation Service.
 U.S. Government Printing Office, Washington, DC.  194 pp.
                                        135

-------
Van Johnson, W.M. and Maxwell, JA (1981). Rock and Mineral Analysis. 2nd. ed. J. Wiley and Sons,
 Inc. New York. 489 pp.

Wentworth, O.K., Wilgus, W.L and Koch, H.L (1934). A Rotary Type of Sample Splitter. J. Sed. Pet.
 4:127-138.
                                         136

-------
                              Appendix E

                Determination of Dust Constituents
              at the Preparation Laboratory Facility
      The following study was initiated to determine if creating a preparation facility in the Las
Vegas area would lead to sample contamination due to the prevalence of many atmospheric dust
constituents that are common to the region.
                                   137

-------
                     Determination of Dust Constituents  at the
                MASS Preparation Laboratory in Las Vegas, Nevada
Introduction
   A proposal was suggested to locate the soil preparation laboratory for the MASS in Las
Vegas, Nevada. Concern was expressed about alkaline dust contamination of acidic soil
samples from the Mid- Appalachian region. Soils in the Las Vegas area are alkaline and have
abundant gypsum (CaS042H20) which could be a source of Ca and SO4 contamination to soils
especially those with low analyte levels (Speck, 1985).


Objectives

   To determine the amount and content of dust at the site of the proposed preparation
laboratory and if the amount present is a source of significant contamination for soil samples
prepared at the laboratory.


Materials and  Methods

   The proposed soil drying area of the warehouse was divided into two areas and dust was
collected to represent typifying dust that has accumulated through multiple weather events over
time.  One bulk dust sample was taken from the north end and one from the south end of the
warehouse.  Both locations were near doorways (Figure 1).

   The samples were passed through a 140-mesh  sieve (0.01 mm) to remove human
contamination  (sawdust, metal shavings, etc.)  and  to remove large organic materials (leaves,
twigs, etc.).  Extractable SO4 was analyzed using 1C and 0.016 M sodium phosphate
(NaH2PO4.H2O) as an extracting solution (Johnson and Todd, 1983).  Total elemental Ca, Mg, K,
Na, Al, Fe, and Mn contents were analyzed on  an ICP after using the bomb method of digestion
with hydrofluoric acid (Bernas, 1968).  The pH was  determined in 1:1 DDI:H20 using a
combination electrode.

   The amount of dust accumulated over a three day period (the average sample drying time)
was also determined prior  to dust control measures implemented during three separate weeks.
Sheets used to collect the  dust were placed throughout the proposed soil drying area were
grouped into three regions, namely, the north, south and middle regions. Vertical fallout  was
collected by laying a 22 sq.in. piece of paper on a flat surface to simulate the condition of an
uncovered sample during air-drying. Lateral fallout  was collected using  two 22 sq.in. sheets of
paper and a 14-3/4 sq.in. by 1-1/2 in. frame  (approximate height of a bulk soil spread to dry).
One sheet was laid flat, the frame was placed in the center, and the second sheet was placed
on top of the frame.  This  procedure simulates the  covering of a bulk soil sample for drying as
outlined in the  preparation  laboratory standard operation procedures (Bartling et al., 1988).

   The sheets were placed on tables in the proposed drying area of the warehouse. After
three days the dust was collected by brushing it to the  center of the paper and transferring it
into a pre-weighed container.  The weight of the dust was determined by difference.

    After dust control measures were implemented at the preparation laboratory, including the
installment of four ceiling fans, and enclosing the soil drying area, twelve (12) uncovered sheets
of paper  (similiar in dimensions to those use in routine  soil drying) were placed randomly on
tables throughout the drying area for thirteen (13) days  in an attempt to collect a measurable
quantity of dust.
                                        138

-------
 Results and Discussion

    Results of the accumulated dust collected over a three-day period for three separate weeks
 are presented in Table 1. The greatest amount of dust was recovered during the first week at
 the south end from lateral fallout. During the first sampling period, a major wind storm with
 gusts exceeding 50 miles per hour occurred and the south warehouse delivery door was opened
 for approximately one hour.  No major wind events occurred during the sampling period for the
 following two weeks and the only samples with a measurable amount of dust were sampled at
 the middle location  with an average total dust collection of 0.015g.

    The pH of the dust was moderately alkaline, ranging from 7.83 to 8.26 (Table 2). Due to the
 trace quantities of the dust collected, the effect of the dust addition on soil pH will be
 negligible.

    Three major assumptions were made for the proposed dust contamination problem to
 determine if the dust, prior to dust control measures being implemented, could cause
 measurable  contamination to Mid- Appalachian soil samples. The assumptions were that:  1)
 the only cation was calcium and the only anion was sulfate, 2) all dust present was gypsum
 (CaSO4.2H20), and 3) all material was in the exchangeable form.


       Table 1.  Weight of Dust  Recovered (Grams) During a Three-day Period
Flow

Lateral
Vertical
Week 1 (location)
So. End
0.10
0.01*
Mid
0.01
0.04
No. End
0.00
0.04
Week 2 (location)
So. End
0.00
0.00
Mid.
0.01
0.01
No. End
0.00
0.00
Week 3 (location)
So. End
0.00
0.00
Mid
0.01
0.03
No. End
0.00
0.00
       •Indicates disturbance to the sample


    Results from total calcium and extractable sulfate were used in the assessment of
contaminant additions because these two elements were found in greatest abundance,
excluding total Si (Table 3), and would thus be of the greatest concern. The following
calculations show that only 0.01785 mg S/Kg and 0.9727 mg Ca/L would be added to each
sample in the prep lab assuming at this point that the sample was entirely composed of Ca
and SO4.  It should be  noted at that the added SO4, on the basis of assumption 1, is less than
the contract required detection limit (CDRL) set for the Mid-Appalachain soil survey and
therefore dust contamination prior to dust control measure implementation poses no problem as
related to contamination of the soil sample. However, the results for Ca are above the CRDL of
0.05 mg/L so Ca is still a concern under the first assumption.
         Table 2. Dust pH Values and Extract Solution Concentrations (mg/L)
Sample ID
South End
Rep 2
Rep 3
MEAN
North End
Rep 2
Rep 3
MEAN
PH
7.83
7T8T3
8.26
8.26
Ca
12.35
13.01
15.02
13.46
17.39
28.07
15.49
20.32
Mg
6.83
6.69
6.91
£.81
6.78
10.47
8.32
8.52
K
3.82
4.12
4.98
4.31
3.11
4.78
4.98
4.29
Na
3.40
3.40
3.70
3.50
2.50
2.80
2.50
2.60
Pe
3.98
4.20
3.87
4.02
2.73
4.02
3.12
3.29
Al
5.28
5.15
5.42
5.28
4.18
«,19
4.88
5.08
Mn
0.089
0.087
0.085
0.087
0.060
0.086
0.075
0.074
S04
852
788
767
802
766
571
722
686
   The second assumption stated that all the dust contaminant was qypsum.  Petrographic
evaluation of the dust revealed that quartz is the dominant mineral comprising 75-80% with
minor constituents of 15- 20% organic matter, and approximately 5% hornblende, biotite,
plagioclase feldspars, gypsum, and olivine.  Silica was the major elemental component and
comprised about 88% of the samples supporting the petrographic mineral analyses (Table 3).
From the results and the data from the first assumption, the following calculation is possible:
                                        139

-------
  Less than 5% of dust is gypsum - 0.09727 mg Ca/L x 5% = 0.0048635 mg Ca/L

   This concentration of Ca (0.048635 mg Ca/L) is less than the CRDL for Ca (0.05 mg/L) set
for the Mid-Appalachain soil survey. Therefore, after correction of  assumption 2, the Ca
content added to the soil  samples from dust contamination prior to dust control measure
implementation poses no  detectable contamination threat to the MASS samples.
         Table 3. Total Oxide Percentage in the Bulk Dust Sample
Sample ID
South End
2
3
Mean
North End
2
3
Mean
CaO
3.46
3.64
4.20
3.76
4.87
7.85
4.34
5.68
MgO
2.27
2.22
2.29
2.26
2.26
3.47
2.75
2.82
K2O
0.32
0.99
1.20
1.04
0.75
1.15
1.20
1.03
Na20
0.92
0.99
1.00
0.94
0.67
0.76
0.67
0.70
Fe203
1.12
1.18
1.09
1.13
0.77
1.13
0.88
0.92
A1203
2.00
1.95
2.04
2.00
1.59
2.34
1.85
1.93
MnO2
0.028
0.027
0.027
0.027
0.019
0.027
0.024
0.024
Si02*
89.28
89.07
88.15
88.84
89.07
83.27
88.29
86.90
        •Calculated by difference

Calculations

The average SO4 concentration in the dust =  744mg/L (Table 2)
    [4g soil and 100ml extract]

    744mg/L * 100ml/4g * 1L/1000ml * 1000g/Kg = 18,600mg SOJKg soil
                                            = 18.60g SO4/Kg soil
                                            - 0.0186g SOJg soil

    0.0002678g SO4 per 0.0144g average collected dust (Table 1)
If a bulk sample = 5kg, then 0.002678g/5000.0144g
          which is 0.05355mg SO«/Kg soil

          To convert mg SCVKg ->  mg S/Kg
            .05355 x .33333 = .01785mg S/Kg
                                                 5.355 x IQ-'g SOjQ soil
The average total elemental Ca concentration = 16.89mg/L (Table 2)
    [O.OSg soil and 100ml extract]

    16.89mg/L * 100ml/0.05g * 1L/1000ml * 1000g/Kg = 33780mg Ca/Kg soil
                                         = 33.78g Ca/Kg soil
                                         = 0.03378g Ca/Kg soil
   0.0004864g Ca per 0.0144g average collected dust (Table 1)

If a bulk sample - 5kg, then 0.0004864g/5000.0144g =  9.727 x

= 0.09727 ppm total Ca > CRDL of .05 ppm exchangeable Ca
                                                               Ca/g soil
    The third assumption made during the analyses was that 100% of the total Ca would
dissolve in water to become exchangeable Ca. According to Brady (1974), only 25% of the total
Ca in soils is in the exchangeable form. Using this percentage reduces the Ca content in the
dust to 0.00122 mg Ca/L which is even further below the CRDL, supporting the conclusion that
the dust contamination potential is well below detectable limits and should not be considered a
problem at the prep lab.

    Results presented represented the situation at the warehouse before dust control measures


                                        140

-------
were applied.  The following major dust control measures reduced the possibility of
contamination:  (1) constructing a free-standing enclosure within the warehouse, (2) sealing off
the south delivery door,  (3) coating walls and floors to prevent the addition of concrete dust,
and (4) sieving and homogenizing in an air-controlled environment. After these dust control
measures were implemented there was no measurable amount of dust collected from the
twelve sheets of uncovered paper in the soil drying area  after thirteen (13) days.

Conclusions

   The results show that the amount of dust recovered  was negligible and that SO4 and Ca
addition were well below the contract-required detection  limits of the instruments to be used in
the survey prior to control measures being enacted. After dust control measures  were
implemented no dust was found collected after a significant length of time, greater than three
days (normal soil drying time); thereby, indicating a lack  of dust for any form form of
contamination of the soil samples. Therefore, there will be no significant addition of any
targeted  element to the  bulk soil sample from dust in the preparation laboratory.
                                         141

-------
                                   References
Bernas, B.  1968.  A New Method for Decomposition and
          Comprehensive Analysis of Silicates by Atomic Absorption
          Spectrometry. Anal. Cham. 40:1682-1686.

Brady, N. C. 1974. The Nature and Properties of soils. Macmillian
          Publishing Co., Inc.

Johnson, D. W.. and D. E. Todd.  1983.  Relationship Among
          Fe, Al, C, and Sulfate in a Variety of Forest Soils.
          Soil Sci. Soc. Am J. 47:792-800.

Speck, R. L 1985. Soil Survey of Las Vegas Valley Area, Part of
          Clark County, Nevada. USDA-SCS.  U.S. Government Printing
          Office Washington, D.C.
                                     142

-------
                                 Appendix F

                Kraft Paper and Nylon Mesh Analysis
      The following appendix contains: (1) a table of data resulting from the analysis of the "PAL"
samples at the MASS analytical laboratories, and (2) an internal report that describes the analysis
of the kraft paper and nylon mesh used for air drying at the preparation laboratory. The table and
report address some data user concerns that there might possibly be contamination leached into
the MASS soil samples from these two materials.
                                        143

-------
Table F-1. Comparison of LAL and PAL Sample Analytical Data from Three Soil Batches
Parameter
SAND
VCOS
COS
MS
FS
VFS
SILT
COSI
FSI
CLAY
PH H20
PH~002M
PH~01M
CA CL
MG~CL
K CL
NA_CL
CA OAC
MG OAC
K OAC
NA_OAC
CEC CL
CEC OAC
AC_BACL
CA CL2
MG CL2
K CL2
NA CL2
FE CL2
AL_CL2
FE PYP
AL~PYP
FE AO
AL AO
FE~CD
AL~CD
SO4 H20
S04~P04
SO4~~0
S04~2
SO4~~4
S04~8
S04 16
S04J32
C TOT
N~TOT
S~TOT
Reporting
units
wt %
II
II
n
it
n
li
II
11
n
pH units
II
II
meq/100g
II
n
n
meq/100g
H
II
"
meq/100g
11
n
meq/100g
"
n
n
11
11
wt %
«
n
"
n
ii
mg S/kg
II
mg S/L
n
n
n
11
"
wt %
II
n
30128
LAL
96.600
6.000
16.200
32.700
31.700
10.000
3.300
3.100
0.200
0.100
5.320
4.850
4.770
0.089
0.031
0.026
0.000
0.115
0.044
0.028
0.011
0.609
1.122
0.210
0.740
0.021
0.015
0.007
0.001
0.005
0.044
0.081
0.052
0.083
0.137
0.052
2.803
3.403
0.480
2.624
4.409
8.492
16.496
32.974
0.089
0.003
0.002
PAL
96.300
4.600
16.200
32.200
33.700
9.600
3.500
3.300
0.200
0.200
5.320
4.870
4.790
0.142
0.039
0.030
0.000
0.136
0.042
0.031
0.015
0.609
1.065
0.889
0.750
0.016
0.013
0.009"
0.000
0.004
0.035
0.080
0.041
0.098
0.210
0.055
2.603
3.003
0.641
2.572
4.438
8.586
16.560
32.479
0.076"
0.004
0.002
30129
LAL
96.900
7.500
15.500
34.100
32.600
7.200
3.100
0.900
2.200*
0.000
5.210
4.980
4.720
0.084
0.032
0.027
0.006
0.054
0.023
0.025
0.006
0.497
0.936
0.201
0.775
0.020
0.006"
0.004
0.001
0.006
0.036
0.056
0.200*
0.092
0.159
0.060
1.802
6.006
0.458
2.458
4.400
8.160
15.700
31.160
0.112
0.006
0.000
PAL
94.700
4.700
13.900
29.100
34.900
12.200
5.500
5.200
0.300
-0.200
5.240
4.950
4.730
0.064
0.025
0.026
0.009
0.073
0.027
0.029
0.010
0.463
0.958
1.126
0.756
0.022
0.014
0.007
0.001
0.005
0.030
0.047
0.057
0.063
0.143
0.050
1.602
5.005
0.431
2.353
4.225
8.080
15.880
31.240
0.094
0.004
0.000
30130
LAL
97.200
4.300
18.00
32.400
33.400
9.000
1.800
1.700
0.100
1.000*
5.460
5.150
4.800
0.057
0.019
0.025
0.006
0.049
0.025
0.020
0.005
0.741
0.741
0.405
0.761
0.028*
0.023*
0.009*
0.002
0.002
0.032
0.035
0.051
0.067
0.116
0.033
2.600
4.800
0.466
2.285
4.186
8.183
16.223
33.771
0.088
0.004
0.001
PAL
97.500
9.700
19.400
30.800
30.000
7.700
1.500
1.300
0.200
1.000*
5.370
5.150
4.750
0.064
0.020
0.027
0.011
0.061
0.022
0.024
0.009
0.741
0.836
0.486
0.410
0.014
0.011
0.024*
0.001
0.000
0.034
0.049
0.058
0.078
0.100
0.059
16.800*
4.000
1.980*
2.704
4.320
8.415
17.058
34.658
0.176
0.005
0.004
NOTE: No accuracy windows were established for MOIST, AL_CL, and SI_AO.
" Value below lower accuracy window limit.
* Value above upper accuracy window limit.
                                                144

-------
                      Internal Report

       DDRP Kraft Paper  and Mesh Analysis
                          by
                     L.K. Fenstermaker
                Environmental Research Center
                University of Nevada-Las Vegas

            Cooperative Agreement No. CR814701-01
                   Analysis performed by
             J. Lindner, V. Letourneau, P. Nowinski,
          S.H. Pia, DA Jadhon, B.A. Schumacher, and
                    R.D. Van Remortel
          Lockheed Engineering and Sciences Company
                    Las Vegas, Nevada

                  Contract No. 68-03-3249
                      Project Officer

                      Louis J. Blume
            Exposure Assessment Research Division
          Environmental Monitoring Systems Laboratory
               Las Vegas, Nevada 89193-3478
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
      OFFICE OF RESEARCH AND DEVELOPMENT
     U.S. ENVIRONMENTAL PROTECTION AGENCY
            LAS VEGAS, NEVADA 89193-3478
                        145

-------
                                     NOTICE
      This document is  intended for internal Agency use only.  Mention of trade names  or
commercial products does not constitute endorsement or recommendation for use.
                                      146

-------
                                   INTRODUCTION
       The work described in this report was initiated as a result of peer comments on the review
of the EMSL-LV activities for the mid-Appalachian Direct Delayed Response Project (DDRP).  The
peer review panel was concerned about the potential for contamination of the mid-Appalachian soil
samples during the drying  process from kraft paper upon which the soil is laid to dry.  External
auditors requested that the mesh table upon which the kraft paper was laid be analyzed as  well.
The Lockheed Engineering and Sciences Company (LESC) preparation laboratory is responsible for
this procedure.  LESC set up a special room in a warehouse for the drying process. The room has
limited access and is vented, and procedures had been established prior to operation to prevent
contamination from dust or other samples.  Drying tables were constructed from nylon mesh and
polyvinyl chloride (PVC) pipe.  At a minimum two sheets of kraft paper (approximately 1  x 1 m) were
placed on top of the mesh for each  sample.  If a sample contained free-standing water, then a
sheet of clean plastic was placed on  top of the paper and the plastic sample bag with the sample
inside was cut open on top of the plastic sheet.  All other samples were placed  directly on the
paper for drying.   In all cases, samples were covered with a sheet of paper to  prevent any
contamination by dust.

       The concern expressed by the peer review panel and external auditors was that, during the
course of drying, there might potentially be movement of ions from  the paper and mesh into the
soil sample. To assess this potential, LESC was asked to analyze  the mesh  and paper for ions
for which the soil  samples would later be analyzed, namely, SO4, Al, Ca, Mg, K, and Na.

       The analytical work described in this report was performed by Lockheed Engineering and
Sciences Company for  the EPA Environmental Monitoring Systems Laboratory (EMSL-LV) under
contract number 68-03-3249, and the report was written by the Environmental Research Center,
University of Nevada-Las Vegas, for EMSL-LV under cooperative agreement number CR814701-01.


                                      METHODS

       Samples of the nylon mesh  and paper used for the drying process  were prepared for
analysis in the following manner. Twenty 4 inch by 4 inch pieces of mesh were cut from the roll
used to make the drying table surface.  Nylon mesh was used for the table surface  to promote
rapid drying of the soil samples.  The mesh had a border along the edge. While this border would
not be located directly beneath the drying soil samples,  it was decided to analyze  mesh samples
with borders to fully examine all possible situations.  Therefore,  half of the twenty  mesh samples
had borders while the other half did not. The typical area of mesh utilized for drying a  soil sample
was approximately 1 x 1 meter (1550 square inches), equivalent to 100 of the 4" by 4" samples. The
twenty samples were placed in 500 ml bottles and brought to volume with double deionized water
(DDI or Type II reagent grade water). The water and mesh remained in the bottle for three days
before separation  and analysis.

       Three  kraft paper samples, each approximately 1 x 1 meter and  weighing 100g on the
average, were placed in one-liter cylinders and one liter of DDI water was added.  After a 24  hour
soaking period the paper and water were separated to prevent dissolution and physical breakdown
of the paper in the water.

       The methods of analysis for both the paper and mesh leachates were ion chromatography
for S04, ICP for Al, Ca, and Mg, and flame atomic absorption spectrometry for K and Na. For the
paper leachates, analyses were run in triplicate except for K and Na which were run in duplicate.
Quality assurance samples consisted of blanks, drift blanks, standard reference samples, and
duplicates.


                                        147

-------
                           RESULTS AND CONCLUSIONS
       Tables 1 and 2 below present the analytical results. To examine a worst-case scenario, it
was assumed that all ions present in the mesh and paper leachate would go directly into the 5 kg
of drying soil.  The maximum possible amount of ion addition to the soil from the paper and mesh
are presented in Table 3 in milliequivalents per 100 grams (meq/100g).  Further calculations were
made to show how much of an increase could be expected during routine soil analysis (Table 3)
and whether this increase would be detectable, i.e., above the  program defined detection limits of
.025 mg/kg for SO^ .1  mg/l for Al,  and .05 mgfl  for Ca, Mg,  K,  and Na.   Only the highest
concentrations  measured, presented in  boldface print in Tables 1 and  2, were used for the
calculations after blank correction. The purpose of Table 3 was to provide two different means of
presenting and understanding what the maximum addition of ions  from the paper and mesh to the
soil would be in common reporting units.  Appendix 1 provides an example of how the calculations
were performed.

       Any potential contamination from the mesh would be detectable for all ions except Al (Table
3).  Even though this potential exists, movement of ions from mesh to soil is severely restricted due
to the absence of water transport from the mesh to the soil.  The only mechanism which would
provide  transport of ions into the  soil would be a diffusion process between the paper and soil
during equilibrium after initial wetting of the paper by soil  water.  Convective movement (or water
transport) of  ions is not a possible transport mechanism as there was not a continuous source
of water.  LESC personnel very meticulously monitored the conditions of the drying soil and the
paper and mesh under the soil. They never noticed any staining  of the bottom layer of paper by
the mesh or any signs of water movement onto  the bottom  sheet of  paper from the soil.
Therefore, diffusion of ions from the mesh to the paper could not occur and contamination from
the mesh to the soil sample  is not possible.

       The potential for  detectable contamination from the paper could occur for SO and Na if a
complete diffusion of the ions from the paper to the soil occurs. While this is not likely, this study
does not provide the necessary information to fully access this. If it is determined that the levels
reported in Table 3 for the extreme case are not acceptable, a follow-up study should be performed.
As suggested previously, the best approach to determine how much the paper is contaminating the
soil is as follows: moisten  low-level audit soils to field moisture and allow them to dry on the
paper in the same manner as the mid-Appalachian soil samples; analyze the samples according
to the DORP  Methods Manual; and examine the results for any  increases that could be due to
contamination from the Kraft paper.  In this manner the possibility of potential contamination can
be accurately assessed under normal routine sample conditions.  Given that saturated soils were
either dried on plastic sheets  or were not allowed extended  contact with the paper, the paper
contamination issue should not be of concern for the mid-Appalachian DDRP soil survey.  However,
extended contact between paper and soil may have occurred  in the previous Northeast and
Southern Blue Ridge survey and therefore justify further research.
                                         148

-------
TABLE 1.  ANALYTE CONCENTRATIONS (MG/L) FOR NYLON MESH
Sample #
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
AVG
STD
Blank- 1
Blank-2
Blank-3
Blank-4
Blank-5
AVG
STD
SO4
0.22
0.47
0.52
0.30
0.24
0.28
0.24
0.21
0.23
0.22
0.20
0.19
0.26
0.35
0.20
0.31
0.32
0.18
0.28
0.26
0.24
0.13
<.01
<.01
0.01
<.01
<.01
__
—
A)
0.041
0.036
0.05
0.04
0.041
0.042
0.045
0.051
0.037
0.04
0.044
0.047
0.034
0.05
0.044
0.047
0.038
0.042
0.036
0.038
0.042
0.01
0.036
0.043
0.047
0.049
0.044
0.040
0.01
Ca
1.07
0.54
0.6
0.41
0.52
0.51
0.6
0.75
0.8
0.78
0.76
0.59
0.58
0.48
0.64
0.36
0.82
0.75
0.81
0.68
0.65
0.17
<.01
<.01
<.01
<.01
<.01
.-
—
Mg
0.088
0.047
0.058
0.042
0.052
0.049
0.049
0.066
0.071
0.067
0.066
0.051
0.049
0.042
0.051
0.037
0.1
0.065
0.067
0.058
0.059
0.02
<.01
<.01
<.01
<.01
<.01
__
—
K
0.172
0.018
0.031
0.034
0.021
0.021
0.023
0.021
0.09
0.032
0.021
0.023
0.026
0.023
0.031
0.013
0.194
0.042
0.058
0.019
0.046
0.05
<.01
<.01
<.01
<.01
<.01
„
—
Na
0.17
0.036
0.056
0.058
0.061
0.05
0.054
0.092
0.078
0.049
0.049
0.04
0.06
0.054
0.059
0.045
0.266
0.075
0.104
0.038
0.075
0.05
<.01
<.01
<.01
<.01
<.01
__
—
                     149

-------
             TABLE 2.ANALYTE CONCENTRATIONS (MG/L) FOR DRYING PAPER
Sample #
Blank
1
2
3
S04
<.13
40.602
35.945
39.024
Al
0.016
0.090
0.091
0.079
Ca
0.008
2.811
3.053
2.616
Mg
0.001
0.298
0.304
0.261
K
<.10
1.34
1.45
1.35
Na
<.30
28.80
28.25
28.00
      Al, Ca, Mg, and SO4 are means of triplicate measurements. Na and K are means of
      duplicate measurements.
       TABLE 3.  ADDITIONS OF ANALYTE TO SOIL (MEQ/100G) AND TO EXTRACTION
Analyte
SO4
Al
Ca
Mg
K
Na
Mesh:
Addition to
soil (meq/100g)

0.0103
0.0010
0.0509
0.0078
0.0047
0.0070
Mesh:
Increase in
soil extraction

1.65 mg/kg
0.0079 mg/l
0.8476 mg/l
0.0792 mg/I
0.1537 mg/l
0.1347 mg/l
Paper:
Addition to
soil (meq/100g)

0.0169
0.0002
0.0030
0.0005
0.0007
0.0250
Paper:
Increase in
soil extraction

2.71 mg/kg
0.0012 mg/l
0.0509 mg/l
0.0050 mg/l
0.0242 mg/l
0.4800 mg/l
                                       150

-------
                                       APPENDIX 1


       Since all analytical measurements were reported for the volume of mesh or volume and
 weight Kraft paper only it is necessary to do some calculations to convert the data into numbers
 which clearly indicate the maximum analyte addition to the soil could be expected if contamination
 occurred.  This appendix provides an example of the calculations which were performed to generate
 Table 3.  The example calculation will use the result from the calcium analysis.

       Knowns:
           *  4 x 4 inches of mesh were leached in 500 ml DDI water and the measurement
               results were reported in mg/l;
           *  1 x 1 meter of Kraft paper were leached in 1 liter DDI water and the measurement
               results were reported in mg/l;
           *  the average area of paper and mesh used for soil drying was 1 x 1 meter;
           *  the average amount of soil per sample  for the mid-Appalachian survey was 5 kg;
           *  and for all analyses done in this study,  except SO4, 5 grams of soil are extracted in
               approximated 60 ml  of extract (SO4 analysis requires 4 grams of soil and 80 ml
               extract).

       Assumptions:
           *  to examine the maximum amount of contamination possible it was assumed that all
               of the ions from the paper and mesh would go into the soil sample.

       The first step is to figure out how much analyte would come from a 1 x 1  meter area of
 mesh from the results for the 4x4 inch mesh sample.  The maximum amount of calcium measured
 in a mesh sample was 1.07 mg/l.  Since 4x4 inches was extracted in 500  ml, two times 4x4
 would be extracted in one liter and therefore 1.07 mg/l Ca is equivalent to 1.07 mg Ca per 2(4 x 4)
 inches.

              mg/l Ca = mg Ca/2(4x4);  total mesh = 1x1 m    =  1521 inch2

              1.07 mo/1 Ca =  x ma/I Ca       = 50.858 Ca mg/l (total mesh)
                32 inch2       1521 inch2

       Assuming all 50.858 mg Ca goes into the 5 kg soil sample, the increase in  Ca to the soil
 would be:

              20.04 mg Ca per milliequivalent
              50.858 mg Ca per 5 kg soil  =   2.538 meq Ca per 5 kg soil
                  20.04 mg/meq

              2.538 meq Ca   =   x meq      =   .0508 meq Ca
                5 kg soil         100 g            100 g soil

       To  calculate the maximum increase to the analytical result that could be expected if 5 g of
the 5 kg soil sample were extracted in 60 ml NH4OAc or NH4CI and all of the Ca in the 5 g sample
originally from the mesh was extracted the following were performed.  (This also points out whether
the increase from the mesh would be above the CRDL's.)

             50.858 mg Ca   =  x mg    =   .0509  mg Ca
               5 kg soil         5 g soil        5 g soil (or 60 ml)

             .0509 mq Ca  =  x mg    =    .8476 Ca mg/l
                 60 ml         1000 ml
                                       151

-------
                              Appendix G

                  Technical Systems Audit Reports
      The following appendix consists of reports that serve to document the technical systems
audits conducted at the Mid-Appalachian Soil Survey preparation laboratory.
                                    152

-------
    PRE-OPERATION AUDIT REPORT
                 for
DDRP SOIL PREPARATION LABORATORY
   MID-APPALACHIAN SOIL SURVEY
     Environmental Research Laboratory.
            Corvallis, Oregon
            September 1, 1988
               153

-------
                         PRE-OPERATION AUDIT REPORT
                                         for
                    DDRP SOIL PREPARATION  LABORATORY
                        MID-APPALACHIAN SOIL SURVEY
INTRODUCTION
       A pre-operation technical systems audit was conducted of the preparation laboratory (prep
lab) facilities for the Mid-Appalachian soil survey on 30 August  1988.  The prep lab is located at
4675 Valley View Drive, Las Vegas, Nevada and is operated by staff from Lockheed Engineering and
Management Services, technical support for the U.S.  Environmental Protection Agency  (EPA),
Environmental Monitoring Systems  Laboratory, Las Vegas, NV (EMSL-LV).

Personnel  Contacted:

Louis Blume                      EPA Project Officer
Michael Papp                     Quality Assurance (QA) Representative
Rick Van Remortel                 Prep Lab Manager
Dave Jadhon                      Assistant Lab Manager
Rob Tidwell                       Shipping/Receiving Manager
Daron Peres                      Lab Tech (part-time)

Additional  prep lab staff will include:

Glenn Merritt                      Lab Tech
Stanton Offemson                 Lab Tech (part-time)
Kim Cutler                        Lab Tech
Joe Parolini                       Lab Tech

Audit Team:

Deborah Coffey                   ERL-C QA Specialist
Jeff Kern                         ERL-C QA Coordinator (Deputy)

Documents received during audit visit:


       o     Revised "Mid-Appalachian  Direct/Delayed Response Project Preparation Laboratory
             Standard Operating Procedures" including Appendix B of this document, "Preparation
             Laboratory Computer Tracking/Verification Procedures" as a separate document. The
             computerized sample tracking system was also demonstrated.

       o     Floor plan of the prep lab facility.

       The cold storage facility used for archiving samples from this and previous soil surveys was
also visited. Las Vegas Ice and Cold Storage Co. (formerly Mr.  Ice) is located on Master Road in
(north) Las Vegas.

       The project is covered by a health and safety plan which has been submitted to project staff
for review.

       Overall, the prep lab is organized and most operations are in place. The only major problem
is the lack of a cooler for sample storage upon receipt. New equipment has been purchased, data
sheets have been generated, standard operating procedures  (SOPs) have been posted, lab staff
have been identified and the lab is nearly operational.   Once lab staff come  on board they will
require training.  Provisions for training have been considered.  Staff already working on the project
were aware of QA concerns and requirements and seemed capable of successfully completing the

                                        154

-------
project.  The sample drying area should not experience problems  with dust contamination.  The
design of the sample drying tables is innovative and is expected to facilitate sample drying. Prep
lab staff should be commended for their effort in preparing the prep lab to receive samples.  An
audit is to be conducted after the prep lab is operational by the EMSL-LV QA Officer, Gene Easterly.
The audit  report from  the second review should be  forwarded to  ERL-C QA staff, the ERL-C
technical director, and the ERL-C sampling task leader. Based on the findings of this audit, it is
expected that no major shortcomings in prep lab operations will be identified.

SUMMARY COMMENTS AND RECOMMENDATIONS


      o      It is recognized by all DDRP project staff that it is a priority that the cold storage
              cooler, to be built in the prep lab, is operational by the time samples begin arriving
              from the field.  The prep lab manager is aware of this  problem and will be working
              to resolve  it to the  best of his ability.

      o      Tests of the Weathershade" material used for the drying tables should be made to
              ensure that the material is inert and does not leach contaminants. Two tests were
              suggested; (1) Place a wet  sheet of white paper on the table to check for visible
              leaching of contaminants from the Weathershade", (2)  Leach the Weathershade" in
              a beaker and analyze leachate for  the  presence of at least cations,  metals, and
              sulfate (ideally all chemical parameters  of interest to DDRP).

      o      Thermometers should  be posted  in the  sample drying area  to  monitor  the
              temperature.  Place thermometers at the level of the drying tables in at least two
              locations.  Document temperatures at least daily, and more frequently if necessary.
              A thermograph  (24-hour or weekly) would  also be appropriate for  documenting
              temperature in the drying area.

      o      Initiate the use of calibration notebooks for all balances used in the project.  The
              notebook should document all calibrations  performed in a format similar to that
              suggested below:

        Date  Expected  wt. Observed wt.  Acceptable? Initials


      o      A trap needs to be added to the compressed air system.  A filter is anticipated to
              be added to this system which should eliminate condensation and oil contamination.
              Alternatively a flask with cotton or absorbent will function as well.

      o      The decision not to seal air vents on the roof should be discussed with ERL-C staff
              before a final decision not to do this is  made.

      o      The final version of the quality assurance project plan (QAPP)  (which includes prep
              lab activities) should include  a signed (by EPA project  officer, EMSL-LV QA officer,
              EPA Technical Director) signature approval page.

      *      A telephone logbook should be initiated  by the lab manager to record  all telephone
              calls concerning sample  receipt, labelling and shipping.  (This was not discussed
              during the  audit and so,  is identified here.)

      o      The samples placed in the archive area should be well-labelled for easy  retrieval.
              Boxes should list batch numbers or other codes that can be used to trace samples.
              Information needed to trace  samples and retrieve them  from cold storage should
              be available in written form, on file at  the  prep lab (or in a sample  custodian's
              office following completion of prep lab activities) and  somewhere at the cold storage
              facility.   Ideally, a master list of all  samples, the batch in which each sample is
              packed, and in which survey the sample was collected should be prepared.


                                         155

-------
              It is the responsibility of ERL-C  staff to determine how long samples  will be
              archived. Sample storage currently costs about $5,000/year. Sample disposal will
              also require funds. If samples are to be accessible to other scientists, funds will
              need to be available to retrieve, package and ship samples upon request.

SUMMARY OF AREAS REVIEWED

       Laboratory  facilities reviewed included:
       Trailer #1 -  pH, clod bulk density analyses
       Trailer #2 - Loss on ignition, air dry determination
       Trailer #3 - Sample preparation area (riffle splitter) and subsample preparation area
       Sample drying
       Sample disaggregation and sieving (rock fragment determination)
       Front office
       Sample receipt/ log-in
       Sample shipping
       Archive - Cold storage
       Computerized Sample tracking/laboratory status program
       Soil description computer entry and data verification system
                                         156

-------
     Ti
      I     UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
     f                  OFFICE OF RESEARCH AND DEVELOPMENT
               ENVIRONMENTAL MONITORING SYSTEMS LABORATORY-LAS VEGAS
                                 P.O. BOX 93478
                           LAS VEGAS. NEVADA 89193-3478
                            (7O2/798-21OO- FTS 545-21OO)
SUBJECT:  Preparation Laboratory Audit Report in Support  of  the  Mid-Appalachian
          Direct/Delayed Response Project of October  6,  1988
FROM:  /  Da*v
      •-T^QA Officer

TO:       Files


     The purpose of this report is to document the major  issues  that  arose as  a
result of an on-site audit of the EMSL-LV soil preparation  facility.   The  audit
was conducted approximately two weeks after initiation  of sample analyses.  The
audit was unannounced and conducted by; the Laboratory  QA Officer,  Mr.  Gene
Easterly; the Nuclear Radiation Division Science Advisor and  QA  Officer, Dr.
Stuart Black; the DDRP Field Sampling QA representative, Jeff Kern; and, the
DDRP EMSL-LV Project Officer, Louis Blume.  Attachment  1 of this report contains
an audit questionnaire which documents the specific details of the  audit.

     The preparation laboratory operations were well  organized and  managed
by an enthusiastic and qualified staff.  We thought that operations were
running very smoothly especially considering  this was only  the second week of
analyses.  There were a few areas of concerns identified, many of which were
already in the process of being rectified.

     After the laboratory tour the laboratory managers  were debriefed regarding
these issues and recommendations.  Areas of concern and recommendations were as
follows:

     a.   Subsampling for various analyses should be  on a composite basis
after samples are mixed instead of taking the complete  subsample from one
location.

     b.   Corrections or cross-outs in logbooks should  not  be obliterated.
They should be crossed out with one line, then initialed.   All entries  should
be made in ink.  In addition, we recommended  the laboratory manager should
review, sign, and date each logbook on a weekly basis.

     c.   Samples received that are very wet  should be  laid out  to  dry along
with the interior plastic bag inverted or opened up.  After the  fine  material
adhering to the bag is dried, it should be added to the sample.
                                      157

-------
     d.   It is acknowledged that the method of breaking up aggregates and
passing soil through the 2 mm sieve is somewhat operationally defined;
nevertheless, we asked the technicians to consider some way of standardizing
this procedure.

     e.   Tolerance limits for blank drift of analytical balances should be
established and controlled for precise weighings.  In addition, servicing
of the balances and scheduled preventive maintenance of the balances and pH
meters should be conducted and assessed.

     f.   Some method of carrying flags or comments about anomalous handling
of samples during sampling, shipment, and preparation should be carried forward
through data verification.

     g.   Sample receipt information should be kept in a logbook, or if loose-
leaf forms are kept, they should be sequentially numbered and be permanently
bound.

     h.   Buffer solutions and QCCS solution should be dated upon opening.

     i.   Ventilation to the grinding area needs Improvement primarily from the
standpoint of technician safety.

     j.   Routine samples should not be placed underneath ceiling fans if It
can be avoided.  Although there Is little probability of contamination due to
the Kraft paper covering the drying soils.

     k.   Organic soils should be air-dried as soon as possible considering
there  is a stronger potential for microbial oxidation of sulfur relative to
mineral soils.

     1.   A minor study should be undertaken to determine if sulfur can be
leached from the Kraft paper onto the soil considering Kraft paper is made
through a sulfite process.

     The bulk density analysis was not observed In its entirety because there
were no ongoing analyses.  Therefore, no comments were made, and this should
be observed In a future audit.  A final report was requested on the dust study
and on whatever action was taken to follow-up on a possible contamination from
the plastic mesh that holds up the soil and Kraft paper.  In summary the review
team found no major problem areas and were quite impressed with the operations.

Attachment
                                       158

-------
                         PREPARATION LABORATORY ON-SITE
                            EVALUATION QUESTIONNAIRE
Date:    10/5/88
Laboratory Manager:  Rick  Van  Remortel
Quality Assurance Auditor:  Gene  Easterly
                            Name
Personnel  Present:  R.  Van  Remortel

                   D.  Jadhon
Evaluation Team:
R.
G.
D.
G.
J.
S.
G.
S.
L.
J.
B.
J.
Kohorst
Satterwhite
Peres
Merri tt
Porolini
Offemson
Name
Easterly
Black
Blume
Kern
Schumacher
Burton
           Title

M.S. Soil , 10 yrs. exp. as
  Soil Scientist
B.S. Geology, 4 yrs. exp.
  Oil Exploration
B.S. Biology, 4 yrs. exp.
  Med Specialist
B.S. Aquat Biology, 8 yrs. exp.
  QA and Biologist
1 yr college, 5 yrs. exp. as
  Warehouse Manager
M.S. Fisheries Biology, 4 yrs.
  Aquatic Field Studies
M.S. Gepscience, 4 yrs. as
  Geohydrologist
B.S. Engineering, 2 yrs. exp.
  Soil Scientist

           Title

Laboratory QA Officer	

Nuclear Radiation Division
  Science Advisor, QA Officer
DDRP EMSL-LV Project Officer
  Soil Scientist
DDRP Field Sampling QA	
  Representative
Soil Scientist	

Preparation Laboratory QA
  Liason
                                       159

-------
                               EQUIPMENT AND  SUPPLY


Are the following  equipments and supplies  available and are  there sufficient
quantities?
              Items
Yes  No
Comments
Logbooks?                                         X
— — —— — — — — — — — •.»»••^«•— — •«— — — »—•« — »»—••••— — •" — «•»"••• — — — — — « — — — — —•-"«••— — — «• — — — «••-• — •— — — * — — ™» — — •

Logsheets?                                        X         Raw data forms


Binders?                                          X
_ — — — — — -••• « V» — — •B-B — — — — — a.^,^™ _ — — — ••••••••»»•••» — »» — »•••• — •••—»•* — — •••««™™ — —• W«_K«H«««

Munsell color book?                               X


Pens-permanent black  ink?                        X


Kraft paper - 36" wide?                          X


Plastic gloves?                                   X
— —_ — w— — — — — «.»»„•• — «P — — — •—*•••••**•«•»•• — •• — — — ^ — — — — — — —"• — ""•••« — •» — »-• — — •"—— — — — — •• — — ••— — •••

Furnance gloves?                                  X


Dust Masks?                                       X         More on order


Gallon paint cans?                                X


Aluminum weighing tins?                          X


Convection oven?                                  X


Lab coats?                                        X


Nalgene bottles - 500 ml,                        X         (and 2-L)
100 ml?                                                    and 2 liter


Carboys - 13 gallons?                            X


Crucibles?                                        X


Evaporating dishes?                               X


Beaker - 1000 ml?                                 X


Tongs?                                            X


Ring stand?                                       X


Thermometers 0-100°C?                            X         Also NBS
                                         160

-------
                        EQUIPMENT AND SUPPLY (CONTINUED)
             Items
Yes  No
Comments
Drying trays - constructed
of PVC and nylon mesh?                          X
Rolling pins?                                   X
Rubber stopper - no. 10 size?                   X
Sieves, 2-mrn and 4.75-mm
mesh?                                           X
Riffle-splitter, Jones-type
with 1.25cm openings?                           X        2
Scale or top-loading
balance?                                        X        2
Analytical balance?                             X        2
pH meters?                                      X        2
Desiccators?                                    X        1 large, custom made
Desiccant?                                      X        2 jars
Hot plate?                                      X
Drying oven?                                    X        Same as convection oven
Muffle Furnance?                                X        2
S Standard weights?                             X        1, 500g
Shipping boxes?                                 X
Packing material?                               X
Strapping tape?                                 X
Mops, buckets,  brooms?                           X
Are there copies of the
laboratory's equipment
inventory lists available?                       X
                                       161

-------
                GENERAL ASSESSMENT OF THE PREPARATION LABORATORY
           Questions
Yes  No
Comments
Is a personal computer                          X
available for data entry?

Are the logsheets organized                     X        Black 3-ring binders
in loose-leaf binders?

Are there hoods and exhaust                     X        For bulk density,
fans where needed and is there                           LOI, and riffle
adequate workspace within hoods?                         splitting

Is there a source of compressed                 X        Air compressor w/
air?                                                     in-line filter

Is there a source of deionized                  X        Carboys delivered from
water?                                                   EMSL-Lab
_ _ _ — W^ — »• _—,— -...•.— ——— — ».«—»•——•»— — — — — — ™— • — -» _« • M • •• V •• «*• W _*—*_•»_ _ _ • «».•..»_•..»«•»• ••—— — —— —
Is there an area designated for                 X        Warehouse area
storing sampling equipment?

Is there a systematic way of                    X
disposing of all waste
i.e., chemicals, paper, etc.?

Are hazardous areas clearly                     X        No specific hazard
posted and is there a fire                               areas, but fire extin-
extinguisher and is it                                   guishers are available
accessible?                                              in every trailer

Are chemicals and other                         X
combustible materials
stored in a safe area?
— — — — ^.^.^ — — — —_— — _„.«. — — — — — — •••.•• — •• — •»•• — ««™™»M»™— •— ^— *• — <•• — •—»•— ••—••-•*«— •»—»«••«•«— — •• — .•••*«»—• — ••••*•*—»•
Are all instruments                             X        Logbooks for all
calibrated and standardize                               improvements in pre-
and in proper working condition?                         ventive maintenance
                                                         is needed

Is the balance checked with a                   X        Daily calibration
class S standard weight before
each use, and is the result
recorded in a logbook?

Is the analytical balance located               X        Observed fluctuation on
away from draft and areas subject                        mettle and sartorious
to rapid temperature changes?                            control  limits for drift
                                                         should be established
                                       162

-------
                         STANDARD OPERATING PROCEDURES
           Questions                           Yes  No         Comments
Does the laboratory have a standard
operating procedures (SOP) manual  and
is it available to each analyst/
technician?
Is the SOP manual followed in
detail?
Does the SOP manual contain
quality control practices?
Does the SOP manual deviate from the
procedures required by this project?
X
X
X
If the SOP manual  does deviate,  are
the deviations documented in written
form?
                                       163

-------
                            ORGANIZATION AND PERSONNEL
            Questions
Yes  No
Comments
 Is  the  laboratory  clean  and  well                 X
 organized?

 Do  personnel assigned to  this  project            X
 have the appropriate educational
 background to  successfully accomplish
 the objectives  of  the program?

 Do  personnel assigned to  this  project            X
 have the appropriate level and type of
 experience to  successfully accomplish
 the objectives  of  this program?

 Is  there adequate  staffing to meet               X        6 Technicians
 the commitments of the project in                         6-day work weeks
 a timely manner?                                          Alternating shifts

 Is  there training  provided for                   X
 preparation laboratory personnel?
 •••^•••*» ••••••••••••••^•• — ^••••.» •••••••••••••••••••••••.•-•.••.•^•aii,^.„„„^ — —,_^__ __ _ -r»— -^ —i
 Are the responsibilities of each                 X
 technician fully outlined?

 Does the laboratory manager have                 X
 his/her own copy of the Field
 Sampling Manual and the Standard
 Operating Procedure?

 Before filling out Form 102,  does                X
 the laboratory manager:
  Review data values on Form 101?
 	—	—--__	-_.._____—__.
  Review raw data  in lab notebooks?              X         Ask  to  review,  sign,
                                                          and  date on  weekly
                                                          basis

  Check for adequate and accurate                X
  ID of QC  samples?

Does the laboratory manager have                 X         101  is  computer-
Forms 101 and 102 on file?                                generated
                                       164

-------
                             SOIL  PREPARATION PROCESS
           Questions
Yes  No
Comments
Is cross-contamination between                   X        Good
samples in the drying area  avoided?

Are there separate workspaces for                X
sample drying and for sample
preparation?

Is there adequate workspace for                  X        Area needs
drying, sieving and crushing of                           ventilation
samples?

Is there adequate workspace for                  X
performing soil analyses?

Are riffle splitters and sieves                  X
cleaned between samples?
^ml^^^tff^^mm^m,.»»,• ••.•••••i-Ti-^r-i— — — — — — ••• — ••.•••••••••••••••••••••••••••••••••••••••••••••••••••••'•••••••••I*
Is drying area removed from                     X
reagent storage?

Is the moisture-content sample                   X
removed In an appropriate manner?
fm j, — — • -m-r nr— ••«•!• m-n-^	— — _ — __ — — — — •• — — — — •• — •• — — •• — — ••«•••••••••••••••• — •••••••••••••••••••••
Is the moisture-content sample                       X
returned to the bulk sample?

Is the field moist pH sample removed             X
in an appropriate manner?

Are rock fragments archived properly             X
after the sieving process?

Are adequate measures taken to                   X        One routine sample was
protect samples from indigenous                           stored directly below
dust contamination?                                       ceiling fan
                                        165

-------
                             COLD STORAGE FACILITY
           Questions
Yes  No
Comments
Is the temperature of the cold storage          X        Recorded on  the  time
facility recorded daily in a logbook?                    wheel and in  logbook

Is there adequate space for the                 X
storage of samples?

Is the unit set-up for easy access to           X
samples and is there a systematic way
of shelving and locating samples?

Is a map or key showing the locations                    N/A Archiving not
of archived samples readily available?                   underway yet

Are sample identifications                      X
permanent and legible?

Are sample numbers cross-referenced             X
with field data and filed?

Are the stored samples tightly                  X
closed?

Are there any open samples stored               X
in the cold storage facility?

Is the cold storage facility                    X
secure?

Is access to the facility limited                   X
to only specified laboratory personnel?
                                        166

-------
                       PREPARATION LABORATORY  PROCEDURES
           Questions
Yes  No
Comments
Is there a digital pH meter and  is               X
it capable of measuring pH to +/- 0.01
pH unit?

Is the combination of pH and reference           X
electrodes high quality, low-sodium  glass?

Are there 50-ml plastic                          X
containers?

Are there disposable                             X
stirrers?

Are there NBS traceable pH buffers               X        Opening  date  is  not
of pH-4.0 and 7.0?                                        labeled  on  bottle

Is the electrode stored properly                 X
when not in use?
•..U1..1— 4i» — -»••••• "T»-i- —r	•••••»•!• 	 _,. — _•_ — •.•.•••••»••-••••••«•• — •»•••••••»•—••••-•••••••••••••••••«•»'•••'•••
Is the open pan balance accurate                 X
to +/- 0.01 g?

Is the convection oven equilbrated               X
at the proper temperature for 24 hours?

Is the desiccator equipped with                  X
desiccant and plate?
                                        167

-------
                                    SUMMARY
           Questions
Yes  No
Comments
Do responses to the evaluation indicate
that project and supervisory personnel
are aware of QA and its application to
the project?
          Very impressed with
          staff experti se and
          enthusiasm
Have responses with respect to QA/QC
aspects of the project been open and
direct?
          Very good
Has a cooperative attitude been
displayed by all  project and
supervisory personnel?
          Some resentment to
          analysis of draft paper
          for sulfate contamina-
          tion
Is the overall  QA adequate to
accomplish the  objectives of
the project?
Are any corrective actions  required?
If so, list in detail  below.
          See report
Comments:
                                       168

-------
             UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                        OFFICE OF RESEARCH AND DEVELOPMENT
               ENVIRONMENTAL MONITORING SYSTEMS LABORATORY-LAS VEGAS
                                 P.O. BOX 93478
                           LAS VEGAS. NEVADA 89193-3478
                            (7O2/798-21 DO • FTS 545-21OO)
                                 /IPR26
SUBJECT:  Preparation Laboratory Audit Report,  in  Support  of the
          Mid-Appalachian Direct/Delayed Response  Project,
          conducted March 16, 1989

From:     David G. Easterly  /W*'<^ &• ^a^ojJU
          EMSL-LV Quality Assurance Officer  °

To:       Audit Team Participants

                        TECHNICAL SYSTEMS AUDIT REPORT
                      OF THE SOIL PREPARATION LABORATORY
        FOR THE DIRECT/DELAYED RESPONSE PROJECT MID-APPALACHIAN  SURVEY

     Location:  4675 Valley  View Drive
     Date: March 16 1989
     Audit Participants:

          Gene Easterly
          Stuart Black

          Dan Heggem
          Mike Papp
          Jeff Kern
EMSL-LV QA Officer
EMSL-LV Radiation Division
  Science Advisor & QA Officer
EMSL-LV BAD QA Officer
LESC QA Operations Supervisor
ERL-C DDRP QA Coordinator
     Preparation Laboratory  Personnel:
          Rick Van Remortel
          Dave Jadhon
          Glenn Merritt
          Joe Parolini
          Bryant Hess
          Daron Peres
Prep Lab Manager
Lab Analyst
Lab Analyst
Lab Analyst
Lab Analyst
Lab Analyst
Purpose:
     The audit was conducted  in  order  to  determine if the preparation
laboratory operations were  being run in a manner consistent with Standard
Operating Procedures, and good laboratory practices.
                                    169

-------
General:

     Vith the exception of clod bulk density analysis, all laboratory pro-
cesses, from sample receipt, analysis, to sample shipment had been completed.
A previous systems audit team (Oct 6) had observed these activities.  There-
fore, this audit was devoted to the observation of the clod bulk density
analysis, examination of rav data, and a review of the items of concern listed
on the previous audit.

Audit:

     A brief review of the sample receipt procedures was conducted.  The
sample receipt sheets were reviewed and a suggestion was made to bind these
sheets in a hard cover binder in numerical order.  It was also suggested that
a second copy of these sheets be made.  The forms that were reviewed were
complete.  The sample storage cooler was examined and the storage and handling
of the samples was explained.  Temperature graphs were examined for fluctu-
ations that might compromise the integrity of the samples but it appeared that
the cooler remained at a constant and acceptable level.  The prep lab manager
then began the review of the clod bulk density procedure.  The following
stages of this procedure were reviewed and demonstrated:

          o  clod drying

          o  clod dipping (saran)

          o  clod muffle furnaces

          o  clod sieving and weighing

             - balance calibration log books reviewed.  It was
               observed that some of  the calibration readings
               were not initialed by  the lab analyst.

          o  clod weighing in water

          o  data review

             - it was recommended that the data be protectively
               bound.

     Although there were no samples in the drying area and most of  the drying
tables had been moved out, the prep lab manager was questioned about concerns
the auditors had on the last visit.   It was established that the prep lab
staff had moved tables that were directly under the air intake fans in the
ceiling  thereby reducing the exposure of samples to possible dust contami-
nation.  The auditors also noticed the set up for venting dust from the
sieving  room.
                                     170

-------
     The lab audit was concluded in the main office area where raw data was
examined.  The audit team reviewed their notes and concerns from the last
visit and drew the following conclusions:

             - The prep lab is functioning very well.

             - The laboratory manager and his analysts have
               adequately demonstrated their knowledge of the
               laboratory methods and good QA practices. Their
               enthusiasm and dedication is commendable.

             - The bulk density procedure appeared to be well
               structured and efficient.

     The following issues and corrective actions identified during the
previous audit have been dealt with:

          o  The lab staff began sub-sampling for pH, LOI and content by
             taking a composite sample from 6 or 7 places from the bulk
             sample rather than from one place.

          o  The lab has been making entries in pen and crossing out with
             one line and initialed.  There were a few corrective entries
             that were not initialed and this will be avoided in the future.

          o  The lab adhered to the recommendation that bulk samples that
             were very wet would be laid out to dry on the plastic sampling
             bag.

          o  The lab attempted  to standardize the method of disaggregating and
             sieving bulk samples.  The  lab manager paid close attention  to
             the soil characterization forms noting highly weathered rock
             fragments that needed particular care in sieving.

          o  The lab set a tolerance limit of IX deviation from the type  "S"
             sample weights when calibrating balances.

          o  The sample receipt forms have not been bound and sequentially
             numbered.  This will be accomplished.

          o  The dust study report was delivered  to  the ERL-C staff.

          o  Reports on Kraft paper and  mesh contamination were delivered  to
             ERL-C.  Further study of  the possibility of contamination  is
             underway.

          o  An attempt was made to dry  wet organic samples as soon as
             possible to avoid  microbial action.  Extra fans were set close  to
             these  sample and they were  turned over more frequently than
             mineral samples.
                                    171

-------
     The following issues or concerns were identified on this audit:

          o  All raw data notebooks should be protectively bound.  A second
             copy should be made.

          o  The lab manager mentioned that the entry system did not have
             enough space for free form notes for some data forms.  The
             manager will check if this problem can be rectified.

          o  All balance calibration entries and data corrections should be
             initialed.

     In summary, the auditors were impressed with the organization of the
preparation laboratory and staff and feel that the operation is completing its
tasks successfully and within the specified quality objectives.

Participants:
Michael Papp, Lockheed
Lou Blume, BAD'
Stuart Black, NRD
Jeff Kern
Northrup Services
Environmental Research Laboratory
200 SW 35th St.
Corvallis, OR 97333
                                    172

-------
                              Appendix H

             Quality Assurance Reanalysis Templates
      The following tables represent the evaluation criteria used to to determine whether a run
of samples for an analytical parameter satisfied the measurement quality objectives.
                                    173

-------
Parameter: PH MP

Field
Audit
Low & Pair
QCCS
Internal
duplicate
duplicate
Manager
sample
MAJOR
If any 2 of 3 relationships occur and 2 minor
flags occur:
1 ) FAL beyond accuracy window
2 ) Mean of FAP or FAO beyond accuracy window
3) FAP of FOA SD > 0.15
Out of expected range
None
None
None
MINOR
If 1 of 3 of the relation-
ships listed in the major
column occurs
none
Routine/duplicate SD > 0.15
Routine/duplicate SD > 0.20
Sample beyond accuracy window
NON
If 0 of 3 of the
relationships
listed in the
major column
occur
within expected
range
Routine /duplicate
SD < 0. 15
Routine/duplicate
SD <0.20
Sample within
accuracy window
NOTE:  Reanalysis would be requested on the occurrence of 1 major or 3 minors



       Appendix H-l.   Quality Assurance  Reanalysis  Template  for  the  Determination of Field pH in Water

-------
  Parameter: OM LOI

Field
Audit
Low & Pair
Internal
duplicate
Field
duplicate
Manager
sample
MAJOR
If any 2 of 3 relationships occur and 1 minor
flags occur:
1) FAL beyond QC control limits
2) Mean of FAP or FAO beyond control limits
3) FAP of FAO RSD > 15%
None
None
None
MINOR
If 1 of 3 of the relation-
ships listed in the major
column occurs
Routine/duplicate RSD > 15%
Routine/duplicate RSD > 20%
Sample beyond control window
NON
If 0 of 3 of the
relationships
listed in the
major column
occur
Routine/duplicate
RSD < 15%
Routine/duplicate
RSD < 20%
Sample within
control window
  NOTE:  Reanalysis would be requested on the occurrence of 1 major or 3 minors



Appendix H-2.  Quality Assurance Reanalysis Template for the Determination of Organic Matter by Loss on Ignition

-------
           Parameter: MOIST P
a>

Field
Audit
Low & Pair
Internal
duplicate
Field
duplicate
Manager
sample
MAJOR
If any 2 of 3 relationships occur and 1 minor
flags occur:
1) FAL beyond control limits
2 ) Mean of FAP or FAO beyond control limits
3) FAP of FAO RSD > 20%
None
None
None
MINOR
If 1 of 3 of the relation-
ships listed in the major
column occurs
Routine /duplicate RSD > 15%
Routine/duplicate RSD > 20%
Sample beyond control window
NON
If 0 of 3 of the
relationships
listed in the
major column
occur
Routine/duplicate
RSD < 15%
Routine/duplicate
RSD < 20%
Sample within
control window
           NOTE:  Reanalysis would be requested on the occurrence of 1 major or 3 minors



                  Appendix H-3.   Quality Assurance  Reanalysis  Template  for the  Determination of  Air-Dry Moisture.

-------
                               Appendix I

                 Laboratory Quality Control Charts
      The following control charts represent the measurement quality samples used to evaluate
accuracy within each run of samples at the MASS preparation laboratory.
                                    177

-------
                     Sample Type=FAP Bw  Variable=PH_MP
   5.4-f
X
a.
"So
'o
   5.2-
   5.0
   4-8
                                                       ACC
                                                      . WARN
                                                      •

                                                     - MEAN
   5.2-
OT
'o
2
1  4.8
    4.6
    4.4
                                                       WARN

                                                       ACC
    20SEP88     150CT88
 Figure 1-2.
                              09NOV88
                                Date
   20SEP88      200CT88     19NOV88      19DEC88      18JAN89
                                Date
Figure M.
                      Sample Type=FAP B  Variable=PH_MP
                                                          ACC
                                                          WARN

                                                          MEAN
                                                          WARN

                                                          ACC
04DEC88      29DEC88
                              178

-------
                      Sample Type=FAL O  Variable=PH_MP
6.2-
6.0-
5.8-
:r
Q.
In 5"6-
O
s
2 5-4'
IZ
5.2-
5.0-
4.8,
•
• ' ACC
. ... . . WARN
•
* • •.. ••*
,„.-» ...... ..... , -. , M£AN
• . '
:
WARN
•
ACC

Figure 1-3.
   4.8T
   4.6-
 CO
'o
 0)
   4.4-1
   4.2-
   200CT88
Figure 1-4.
   20SEP88      200CT88     19NOV88      19DEC88     18JAN89
                                Date
                       Sample Type=MS  Varioble=PH_MP
                                    ACC
                                    WARN


                                    MEAN



                                    WARN

                                    ACC
14NOV88
09DEC88
  Date
03JAN89      28JAN89
                              179

-------
                     Sample Type=FAP Bw  Variable=OM_LOI
    8-1
  
-------
                      Sample Type=FAL C  Variable=OM_LOI
CD .5 -
-+-*
-t—'

o



y





i_

o


c


o
i_

CD  '

CL '-
    .2-
Figure 1-7.
    11 -r
    10-
  C

  O
  en
  ^

  O
  c

  CD

  O
  i^_

  CD

  CL
   9-
     8-
     7-
     6-
                                 Date
                       Sample Type=MS  Variable=OM_LOI
   200CT88
Figure 1-8.
               14NOV88      09DEC88


                               Date
                                                          CONT




                                                          WARN







                                                         •


                                                         - MEAN









                                                          WARN




                                                          CONT
          	,	,	,	,	,	!	,



   20SEP88      200CT88      19NOV88      19DEC88      18JAN89
                                                            CONT




                                                           * WARN
                                                          MEAN








                                                          WARN




                                                          CONT
03JAN89      28JAN89
                                181

-------
                     Sample Type=FAP Bw  Vanable=MOIST_P
   2.5-
   2.0-
22 1.5-1
 a>
L_



2 1.0J
   0.5.
   o.o.
                             !    *
Figure 1-9.
                      Sample Type=FAP B  Variable=MOIST_P
 O

Q_
   4.0-
   3.0-
   2.0-
    1.0-
   0.0-
   20SEP88
Figure 1-10
                                                           ACC
                                                          - MEAh
   20SEP88      200CT88      19NOV88     19DEC88     18JAN89


                                 Date
                                              :ONT



                                              VARN








                                              vIEAN
                                              \CC
                                              /VARN



                                              :ONT
150CT88     09NOV88     04DEC88      29DEC88


                Date
                               182

-------
                        Sample Type=FAL C  Variable=MOIST_P
o.u-
2.5-
^ 2.0-
en
"o
•o 1.5-
«
L_
2 1.0-
0.5-

0.0-

ACC







.... «-^.-r,.—i-._. •_. 	 l_...s_.LT. 	 -,-— : , ^ . MEA^

    20SEP88      200CT88

Figure 1-11.
                         Sample Type=MS  Variable=MOIST_P
Figure M2.
9NOV88      19DEC88      18JAN89
 Date
o.u-
4.0-
i — '
| 3.0-
_~0)
i 2.0-
o
Q_

1.0-
0.0-
200

CONT
WARN
. * . . . . *
	 ' ' ACC ,
• • . - MEAN
•
• WARN
CONT


1 i | i | i 	 1 	 1 	 jJ
3T88 14NOV88 09DEC88 03JAN89 28JA
Date
                                 183
                          U.S. GOVERNMENT PRINTING OFFICE: 1990—7 ^ 8 - 15 9/ 20"*59

-------