United States
Environmental Protection
Agency
Environmental Monitoring and
Support Laboratory
Cincinnati OH 45268
Research and Development
EPA/600/S4-86/043 March 1987
Project Summary
Development of Preservation
T— hniques and Establishment  of
                                   Times:
                                    its of  the
                                   )ischarge
Elimination bybio,.i and Safe
Drinking  Water Act
Huge S. Prentice and Daniel F. Bender
  This project was designed to statis-
tically determine the length of time a
sample can be stored for analysis of ten
National Pollutant Discharge Elimina-
tion System (NPDES) compliance pa-
rameters and two Safe Drinking Water
Act (SDWA)  parameters. Accepted
preservation and storage  techniques
were used on spiked representative
real-world samples.  Experimentally
determined maximum holding times
(MHT) were calculated for each pa-
rameter in three different water matrices
and under several different preservation
conditions for certain parameters.
  Fifteen experiments were conducted
for the ten NPDES compliance parame-
ters and three experiments for the two
SDWA compliance parameters. The ten
NPDES parameters  were phenols,
cyanide, mercury, ammonia, nitrate plus
nitrite, fluoride, total Kjeldahl nitrogen,
total phosphorus, total organic carbon,
and sulfide. Multiple experiments were
performed to evaluate MHTs for dif-
ferent  preservation  techniques for
phenols and mercury. The two SDWA
parameters were nitrate and fluoride,
with two experiments performed to
evaluate MHTs for different preserva-
tion techniques for nitrate.
  This Project Summary was developed
by EPA's Environmental Monitoring and
Support Laboratory, Cincinnati, OH, to
announce key findings of the research
project that Is fully documented In a
separate report of the same  title (see
Project Report ordering Information at
back).


Experimental Design and
Approach
  The experiments performed during this
study are summarized in Tables 1 and 2.
For each of the 15 wastewater experi-
ments, three different wastewater sam-
ples were collected. For the three drinking
water experiments, three finished drink-
ing waters were collected from potable
water treatment plants treating different
raw waters using different methods. For
wastewater or drinking water parameters
with multiple experiments (e.g., phenols
and mercury in wastewater and nitrate in
drinking water), sample matrices were
collected at the same site, and mixing
and analyses were started on  the same
day. The protocol to prepare samples for
a time-sequenced  analysis of one pa-
rameter in a matrix was performed 54
times during the project (three matrices

-------
Table 1. Summary of Experimentally Determined MHTs for Wastewater Parameter Experiments
Experiment
number
1
2

3
4
5
6
7
8
9

10
11
12


13
14

15

Parameter
Phenols
Phenols

Phenols
Cyanide
Mercury
Mercury
Mercury
Mercury
Ammonia

Fluoride
TKN
Nitrite
plus
nitrate
Organic
carbon
Total
phosphorus
Sutfide

Preservation
technique
Cool4°C
Cool4°C
1 g/L CuS04
H3P04pH<4
Cool4°C
Cool4°C
NaOHpH>12
HN03pH<2
0.05%K2Cr20J
HNO3pH<2
0.06% Kfr^Oj
HNO3pH<2
HNO3pH<2
Cool4°C
Conc.H,SO4pH<2
None
Cool4°C
Cool4°C
Cone. HzS04pH<2

Cool4°C
Cone. W«SO
-------
for 18 experiments). The protocol for pre-
paring  samples  for  time-sequenced
analysis included:

  1. Collecting grab samples and ship-
     ping  to  the  laboratory within 24
     hours;
  2. Compositing and homogenizing the
     grab  samples in  a large-volume
     mixing container;
  3. Collecting three aliquots for analysis
     from the composite prior to addition
     of preservative;
  4. Collecting three aliquots for analysis
     from  the composite containing the
     preservative;
  5. Adding the  study analyte to the
     preserved composite and mixing;
  6. Collecting 100 aliquots (dispensed
     to ensure a large enough population
     for representative random sampling)
     from  the  spiked, preserved compo-
     site; and
  7. Randomly selecting 18 fixed aliquots
     of the 100 aliquots chosen for time-
     sequences analysis.
The time sequence was either 0, 1, 2, 4,
8, and 16  days or 0, 2, 4, 8, 16, and 32
days.
  The time-sequenced experimental data
were used to calculate experimentally
determined maximum  holding times
(MHTs)  according  to a mathematically
acceptable definition. The experimentally
determined or calculated MHT was de-
fined as the 90 percent lower confidence
limit of a specified critical time (CT) (see
Figure 1). The CT was defined as the time
at which  a  10 percent change  in the
analyte concentration (from the intercept
at Day 0) occurred and when the precision
of the method allowed that 10 percent
change  to be a statistically significant
difference at the 90 percent confidence
level. When the precision of the method
was not sufficient to discern a 10 percent
change in the analyte concentration, the
CT was that time at which the percent
change (greater than 10 percent) in the
analyte concentration represented a
statistically significant difference at the
90 percent confidence level. An upper
limit of a 15 percent change in the analyte
concentration was set for determining a
valid CT. The experimentally determined
MHTs were specified  as  either the 90
percent lower confidence limit (LCL) of
the CT  or the maximum time of the
designed experiment when the LCL was
longer.
  A statistical protocol was developed to
provide information to support the use of
a zero or first order  kinetic model to
calculate a CT and its 90 percent LCL (or
    1.1Cl
    0.9CI
                                                            1O% Gain
                                          Initial Concentration
                            Best-Fit Line of Experimental Data
                            MHT (LCL)    CT

                                     Time

Figure 1.   Graphic representation of MHT definition.
 MHT). Only the raw data zero-order model
 (concentration versus  time)  and  the
 transformed on first-order model (natural
 log of concentration versus time) were
 evaluated since loss or gain of the con-
 stituent was expected to follow zero-or-
 first-order kinetics.

 Results
  A summary of the results  for  the
 wastewater  parameter  experiments is
 presented in Table 1. The experimentally
 determined  MHTs  resulting from  this
 study were longer than proposed MHT
 values published in the Federal Register
 (44/244:75-28-75052, Tuesday, Decem-
 ber 18, 1979) and recommended by EPA
 in Methods  for Chemical Analysis of
 Water and Wastes, EPA-600/4-79-020,
 for the following parameters in waste-
 water:
  Phenols         Nitrate plus nitrite
  Ammonia       Total organic carbon
  Fluoride         Total phosphorus
  Total Kjeldahl    Sulfide
    Nitrogen (TKN)
 A complete set of mercury and cyanide
 experimental results were unobtainable
 because of difficulties with the analytical
 procedures.  The  experimentally deter-
 mined MHTs for fluoride and nitrate in
 drinking water were longer than the EPA
 required  values and  are summarized in
Table 2. Additional analyses for the nitrate
in drinking water experiments supported
experimentally determined MHTs of 32
days.


Conclusions
  From  the  results obtained under the
set of conditions used for this study, the
following conclusions are drawn:
  1.  The experimentally determined
     MHTs were longer than the proposed
     holding times specified  in the
     Federal Register and  the.fitcom-
     mended values from EPA for the
     analysis of the parameters in waste-
     water samples listed below.
      Phenols   Nitrate plus nitrite
      Ammonia  Total organic carbon
      Fluoride   Total phosphorus
      TKN       Sulfide
     For mercury and cyanide, the experi-
     mentally determined MHTs were not
     longer than the proposed or recom-
     mended values.
  2.  The experimentally determined
     MHTs  were  longer than the EPA
     required holding time values for the
     analysis of  nitrate  and fluoride  in
     samples  from  finished drinking
     water.
  3.  Cyanide concentrations in waste-
     water   samples containing  sulfur

-------
   compounds, which are converted to
   sulfide during the distillation proce-
   dure,  cannot  accurately be deter-
   mined using the currently approved
   analytical method for cyanide.
4. The major reasons  that some ex-
   perimentally determined MHTs were
   lower than the proposed and recom-
   mended values were:  the known
   variability for mercury analyses,
   especially at sample concentratons
   less than 2.0 micrograms per liter
   (jug/L), and the possible effective-
   ness  of  0.05 percent potassium
   dichromate (K2Cr207) in samples
   acidified with nitric  acid (HN03) to
   extract mercury from the walls of
   new,  HNOa-rinsed polyethylene
   Hedwin  Cubitainers®*. The slopes
   of the best-fit  regression lines were
   not statistically similar for the two
   experiments  using  polyethylene
   Cubitainers®. Decreasing concentra-
   tions with the time were observed
   when samples were preserved with
   HN03, except at the highest mercury
   concentration where no statistical
   change with time was observed. In-
   creasing  concentrations with  time
   were  observed when 0.05 percent
   K2Cr207 and  HN03  were used  as
   preservative.
5. Different sample preservation tech-
   niques were  studied for nitrate,
   phenols, and mercury analyses. The
   two  preservation techniques for
   finished drinking water for nitrate
   analysis yielded  the same experi-
   mentally determined MHTs. Of the
   three preservation experiments for
   phenols in wastewater, the two in
   which H2SO4 was added to samples
   stored in plastic or  glass provided
   the longest experimentally deter-
   mined MHTs.  Of the four mercury-
   in-wastewater preservation experi-
   ments, HNO3  and K2Cr207 added to
   samples stored in glass provided the
   best  experimentally determined
   MHTs.
6. The phenols experiments that used
   H2S04 as the preservative consis-
   tently exhibited higher coefficients
   of  variation  for the  unexplained
   variability from regression diagnos-
   tics compared to the copper sulfate/
   phosphoric acid preservation experi-
   ments. The highest  CVs were from
   Sample Matrix 3, which also had
   the highest suspended solids con-
   tent of the sample matrices collected
   for the phenols experiments. Since
   the higher variability is due to the
   high suspended solids content, the
   higher CVs were accepted as rea-
   sonable  for analytical variability,
   therefore,  allowing the calculated
   MHT to be considered valid for that
   matrix.
7. A statistically significant  lack of
   linear fit was encountered for the
   experimental data due primarily to
   larger day-to-day analytical vari-
   ability  compared  to  within-day
   analytical variability. This phenom-
   enon is not entirely unexpected, and
   it has caused some concern from a
   purely statistical standpoint. Specifi-
   cally, the concern arises from de-
   riving a precise procedure to evaluate
   when the experimental data exhibits
   a true lack of linear fit. A true lack of
   linear fit for the experimental data
   implies that an  MHT cannot be
   validly calculated.  The procedure
used for  this study represents a
reasonable approach for the deter-
mination of MHTs. Further holding
time studies should recognize the
problems  encountered during this
study and possibly incorporate the
following  two recommendations.
First,  studies should  provide  a
mechanism to precisely define pro-
ject-specific analytical variability for
comparison to lack-of-fit variability.
Second, studies should use  an
analysis of the residuals from the
linear regression line to detect out-
liers and  determine the influence
from extreme observations.
Mention of trademarks or commercial pro-
ducts does not constitute endorsement or
recommendation for use.
 H. S. Prentice. J. E. Singley. L J. Bilello, J. T. McClave. K. L Tuttle. E. M. Kellar. and
   M.  G. Schultz are  with Environmental Science  and Engineering, Inc.,
   Gainesville, FL 32602.
 Daniel F. Bender is the EPA Project Officer (see below).
 The complete  report, entitled "Development of Preservation  Techniques and
   Establishment of Maximum Holding Times: Inorganic Constituents  of the
   National Pollutant Discharge Elimination System and Safe Drinking Water
   Act," (Order No. PB 87-132 833/AS; Cost: $18.95, subject to change) will be
   available only from:
         National Technical Information Service
         5285 Port Royal Road
         Springfield, VA 22161
         Telephone: 703-487-4650
 The EPA Project Officer can be contacted at:
         Environmental Monitoring and Support Laboratory
         U.S. Environmental Protection Agency
         Cincinnati, OH 45268

-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
     BULK RATE
POSTAGE & FEES PAID
        EPA
  PERMIT No G-35
Official Business
Penalty for Private Use S300

EPA/600/S4-86/043
                     0000319    PS

-------