PBS4-1S6089
Capture-and-Containment Systems  for
Hazardous Material Spills'on  Land
MSA Research Corp., Evans City,  PA
Prepared for

Municipal Environmental Research  Lab,
Cincinnati, OH
Apr 84
         of Cesnstwrce
   T^ckitsl ferfwmstkm

-------
                                                     r- L- t: 4 -116 ••: i j

                                                  EPA-600/2-84-084
                                                 -April 1984
         CAPTUPE-AfiD-CONTAT?,'MtNT  SYSTEMS FOR
           HAZARDOUS  MATERIAL  SPILLS  ON  LAND
                           by

                   Mervin 0. Marshall
                MSA Research Corporation
             Evans City, Pennsylvania  16033
                Contract Nc. 63-03-250?
                    Project Officer

                    John t. Bruoqer
       Ofl and Hazardous Materials Spills
Municipal Envlrcmf^ertal Research Laboratory (Cincinnati)
                Edison, New Jersey 08fi37
      MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
           OFFICE OF RESEARCH AND DEVELOPMENT
          U.S. ENVIRONMENTAL PROTECTION AGENCY
                 CINCINNATI, OHIO 452C3

-------
                                  TECHNICAL REPORT DATA
                           (Picas* Tfdd /nu/uctiuns on tne rciene before co
                       ZE:
  I ij O R1 N O
                                                         3 RCCtPlf NT'S
 FPA-6QQ/2-84-Q84
 1 15 Lt AMD SU8TM Lt
 CAPTURE-AND-CONTAINMENT SYSTEMS FOR
 HAZARDOUS MATERIAL SPILLS ON LAND

 Al.iTHORlS>~~
 Mervin  D.  Marshall
                                                          5 REPORT DATE

                                                            _Anril 1984.
                                                         6. PERFORMING ORGANIZATION CODE
                                                         B. PERFORMING ORGANIZATION REPORT NO.
                                                                              8608  9
 PERFORMING ORGANIZATION NAME AND ADDRESS
MSA Research Corporation
Evans City, PA 16033
                                                          1O. PROGRAM ELEMENT NO.

                                                            CBRD1A
                                                          tl. CONTRACT/GRANT NO.

                                                            68-03-2507
12 SPONSORING »GI NCV NAME AND ADOH6SS
Municipal  Environmental  Researc" Laboratory—Gin.,, OH
Office  of  Research and Development
U.S.  Environmental Protection Agency
Cincinnati,  Ori  45268
                                                          II. TVPE OF REPORT AND ft «IOD COVERS O
                                                           Fir-al - Jan. 1977-April  1QRQ
                                                          14. SPONSORING AGSWCV CODE

                                                           EPA/600/14
15 SUPPLEMENTARY NOTES
Project  Officer:   John E.  Brugger (201) 321-6634
16 *BSlW£ report  covers the investigation of methods for  sealing  the  surface of soils an
preventing the percolation of spilled hazardous materials  into the ground.   The objecti
was  to  develop a portable, self-contained, universal sealing  system•"which could be
operated by one  man,  retain the spilled iraterial  through a  24-hr removal  period, and
pose no subsequent hazard to the user or to the environment.
     Sprayable sealant systems were investigated extensively but  proved to be imprac-
tical.   Of the sealants investigated, only polyurethane  foam  showed  any promise of
sealing soil  surfaces particularly under the  severe weather conditions that generally
accompany bulk transport accidents.  However, the quant;ty  of chemical required to seal
soil  (especially grassy substrates) far exceeded  the amount considered practical for a
portable system.
     Polyethylene film in the shape of a one-end-closed tube proved to be the most
practical means  of containing and collecting  hazardous material  spills.  An apron
attached to the  open  end was equipped with draw and tie  ropes to position the device
under or close to the spill.  A small tube connected at  the closed end allowed for
transferring  collected spilled material to auxiliary containers. A  double-walled bag
consisting of a  heat-sealed inner bag and a reinforced outer  bag appeared to be the
combination most suitable for further development.  Total  collapsed  volume of the final
package, a pillow-shaped containment bag 20 ft  long and  6  ft  wide, was less then 2
ft-*,  the  unit M
                           -ttr
  10 lU.
KEV WORDS AND DOCUMENT ANALYSIS
17.
                 DESCRIPTORS
                                             li.lDFNTIFIEnS/OPf N ENDED TERMS
                                                                        c. COSATI Ficid.'Croup
1* OlSTRiBUTlON STATEMENI

RELEASE  TO PUBLIC
                                                                        31 NO OF PAGES
                                                                             79
                                              70 «
                                                St Cu«UT v c L* S5 (Tftit part!
                                                UNCLASSIFJtD
                                                                        72. PRICE
   Form
           |»-7J|

-------
                                 PJ sci. AIMER
The information in this document das  been  funded wholly or  in  part  by the
United States Environmental Protection Agency  under Co-tract  No.
68-03-?b07 to MSM Kesearcn Corporation,   It  has been  subject  to  the
Agency's peer and acministrative review,  and it has foetii  approved  for
publication as an EP* document.  Mention  of  trade  names or  commercial
products does not constitute  endorsement  or  reco(w:en(Jat':on  for use.

-------
                                  FOREWORD
    The U.S. Environmental Protection Agency was  created  because  of  in-
creasing DuL)lic and government concern  about the  rijnqers  of  pollutior. to
the health arid welfare of the American  people.  Nc/ious  air,  foul  water,
red  films such  as  polyethylene, en the other hand,  appear to De
D>-Jctit;dl  and  **ere developed  furtner into a promising, first-action  cap-
wure-and-coiitainment Dag.
                                        Francis T. Mayo,
                                        Director
                                        Municipal Environmental  Research
                                          Laboratory
                                      Hi

-------
                                  ABSTRACT
    This report investigates rethods for sealir.g the surface of soils and
preventing the intrusion of hazardous materials into ground water.  The
objective ^as to develop a portable, self-contained, universal sealing
system that could be operated by one man, retain the spi'lleo material
through a ^4-hour removal period, and pose no subsequent hazar.1 to the
user or to the environment,  K basic asMjirption of this  investigation was
that the site preparation for any seal inn system would be minimal.

    Sprayable sealant systems were  investigated extensively but proved to
be impractical.  Three types cf sealants wore investigated:  nonreactive,
(which are dispersed or dissolved in a solvent system),  reactive  (which
are two or mere co'noonent systems)  and repellent or surface modifiers.
Only polyuretiiane foam, a reactive  type, showeo any promise of sealing
soil surfaces under the variety and severity of weather  conditions that
are often ttit> cause of bulk  transport accidents,   but  even with this
pnvrise, the Quantity of chemical required  to seal soil  (and especially
grassy substrates) sufficiently to  contain  the spilled material far
exceeded the amount considered practical for a portable  system.   In
addition, the xechnn^ues  for  appiyinq the  sealant  to different soil
substrates tinner various  weather  conditions were a biq factor  in
eft-.'Ctinq 3  seal ana thus made  it too complex  for  a simple,
first-resoonse action.

    'olyethylenr film proved  to  he  a  practical barrier 1o  a wide  spectrum
of hazardous ^e;its and  sealed  against water at over a foot of head  when
applied  to rocicy, crassy, and weedy areas.  Mien formed  in  the shape  of  a
tube approximately b ft  in  diameter and  approximately  20 ft  long, the
unit would contain water  with more  than  3  ft of head and have  a  capacity
of more  than 1000 qal.   An  apron  on the  open end equipped  with draw  and
tie ropes allowed positioning of  the  unit  below the  leak,  and  a  small
tube connected at the closed end  allowed for directing excess  spill  to a
second  containment site.  The enclosed bag concept would also  minimize
vapor hazards.

    A double-walled bag  consisting  of a  heat-sealed  inner  bag  of clear
polyethylene and a sewn,  reinforced outer  bag  appeared to  be  the combi-
nation trost  suitable for furtner  development.   The final package had a
volume  of less  than ?. ft3 and weighed  less than  16 Ib.

    This report was submitted  in  fulfillment of Contract No.  68-03-2507
by KSA  Research  Corporation under the sponsorship  o*  the U.S.  Environ-
mental  Protoction Agency.   This  report  covers  the  period
January  7,  19/7  to April  /, 1980.

                                     iv

-------
                                  CONTENTS
Foreword	   iii
Abstract	    iv
Figures	    vi
Tables	   vii

    1.  Introduction ... 	 .....      1
    2.  Conclusions	  .      4
    3.  kecommenaations   	 .....      6
    4.  Modular Capture-and-Containment-Bag Concept	      8
         Concept development  	      8
         Prototype testing 	      9
              Seal strength of conme.-cial P£  film bags	      9
              Testing of  laboratory-fabricated clear  PE  bags  ....    12
              Testing of  f iber-rei'iforced PE  baas	    14
              Testing of  fiber-reinforced bags  from commercial
                   fabricators	  .    15
              Testing of  double-walled  bags  from commercial
                   fabricators.  .	    20
              Summary of  capti:re-and-coTitainment-naq  concept  testing    21
              Field  demonstration	    23

Appendix A -  Investigation of  Surface  Sealing  Systems  	    32
Appendix 8 -  English  to Metric Conversion Factors 	  ....     70

-------
                                  FIGURES
Number                                                                Page

  1      Capture-and-containment device, showing dual-layer
           polyethylene film construction and collecting aoron	 10
  2      Deployment of device as used in simulated field test 	 10
  3      Apparatus used to determine the strength and integrity
           of polyethylene bags	11
  4      Capture-and-containment device with sewn seams being
           tested to maximum containment pressure head	16
  5      Rupture of the device in the transfer tube attachment area  .  . 16
  6      Capture-and-containment device fabricated of "Griff-Weave 24". 19
  7      Capture-and-containment device fabricated by Sheldahl Corp  .  . 21
  8      Construction details of a quarter-scale test device fab-
           ricated by Silco  Industries	22
  9      Quarter-scale capture-and-containment device fabricated
           by Silco Industries	   22
 10      Simulated tanker  accident for  field test demonstration  ...   24
 11      Detail of grommets  for attaching  apron	    25
 12      Detail of the transfer tube  attachment	   25
 13      Weight,  size, and total volume of  these devices  are
           easily handled  by one man	   26
 14      Field demonstration test  - deploying  the capture-and-
           containment device	   28
 15      Field demonstration test  - securing the device to the  tanker   28
 16      Field demonstration test  - detail  of  the apron,  secured to
           collect leakage around  the hatch 	   29
 17      Field demonstration test  - Silco  Industries  device con-
           taining about 200 gal	   29
 18      Field demonstration test  - Silco  Industries  device con-
           taining about 500 gal	,	   30
 19      Field demonstration test  - modular concept  demonstrated.  .  .    30
 20      Overview of  the field  demonstration  site with  the two
           capture-and-containment devices deployed  	    31
 21      Detail of the Katz  Bag Company device containing about
           500 gal	    31
 A-l      Schematic of spray chamber  	   47
 A-2      Compatibility test unit	    43
 A-3      Air  pressurization  apparatus used to  screen test specimens  .    59
 A-4      Modified test apparatus	    66
                                     vi

-------
                                   TABLES
Number,

1        Results of Strength Tests with Commercial Polyethylene
           Bags	13
2        Summary of Tests on Fiber-Reinforced Bags From Commercial
           Bag Manufacturers	  .  17
A-l      Total Sealant Weight as a Function of Application Rate
           and Sealant Thickness 	  35
A-2      Listing of Sealant Candidate Manufacturers	38
A-3      Survey of Sealing Materials 	  41
A-4      Primary Representative Hazardous Chemicals	44
A-5      Results of Compatibility-Permeability Tests with  Borden
           Chemical Company Latex  "Polyco 2607"	52
A-6      Test Results Obtained From Samples of Cliridrol  "Superamide
           lOOCb"	53
A-7      Results of Compatibility-Permeability Tests with  £.  I. duPont
           DeNemours  Surfactant "Zonyl  RP"	54
A-8      Results of Compatibility-Permeability Tests with  Callery
           Chemical i ompany Urethane "115"  Formulation	55
A-9      Results of Compatibilty-Permeability Tests  with Ashland
           Chemical Company "EP 65-36/88".  .  .	56
A-10     Selected  Properties of Three Allied  Chemical  Company
           Polyethylene  Polymers  	  57
A-11     Low-Temperature Curing of Borden Chemical Company
           "Polyco 2607" 	  59
A-12     Low-Temperature Curing of Ashland  Chemical  Company
           Urethane Foam "EP 65-86/88"	59
A-13     Low-Temperature Curing of Callery  Chemical  Company
           Urethane  "115" Formulation   	  60
A-14     Preliminary  Estimates of  Cylinder  Sizes for Supplying
           Atomizing  Air	63
A-15     Calculated Power Requirements  and  Battery Weights for
           Liquid  Coating Applicator Systems  	  63
A-16     Results of Pilot-Scale Tests  Using Urethane Sealants	67
                                     vii

-------
                                  SECTION 1

                                 INTRODUCTION
    Concern about the control of chemical spills has increased in both the
public sector and the scientific community.  The incidence of accidental
land spills has increased proportionally to the growing use and transport
of chemicals.  This concern is at least partly the result of several
incidents in which ground water has been contaminated by percolation of the
spilled material.

    Most of the scientific activity in this area has been directed to
monitoring the fate of the spilled material or health effects resulting
from its presence.  Relatively  little has been done to develop methods to
alleviate the problem of contaminated ground water.  Present action  in-
volves either the physical removal of the soil from the spill-affected area
and disposing of it at a secured  landfill, or  the  drilling of "cone  of
influence" wells to drain off the polluted water for conventional
treatment.  These techniques are  slow and expensive.

    This report contains the results of  an  investigation of methods  to  seal
the surface of soils and prevent  the  intrusion of  hazardous materials  into
the ground.  The objective was  to develop  a portable,  self-contained,
universal  sealing system that could be operated  by one man,  retain  the
spilled material throughout  a 24-hr removal period,  and  pose  no  subsequent
hazard to  the user or  to the environment.

    The hypothetical scenario for the use  of  a soil  sealing  system  assumed
that  it would be employed  immediately  after a spill,  possibly in
conjunction with diking materials,  to confine the  material  to the immedi-
ate vicinity of  the  accident.   The  "esponse procedure  wguld  utilize the
transportation personnel or  local safety personnel who were  first on the
scene  to determine the probable course  of  flow,  to locate  a  containment
site,  and  to  initiate  eapture-and-containntent procedures.

    One  basic  assumption of  this  investigation has been  that site
preparation  for  any  containment or  sealing system would  be minimal.  Con-
ceivably,  some intermediate-sized debris could be moved,  but the ideal
sealing  system would be capable of  surmounting most obstacles.   Since
accidents  frequently occur  during inclement weather, wet or frozen  sub-
strates  have been  considered common conditions for operation.  The design
goals  for  the  system were  as follows.   The system must be:

-------
         1.   Portable and capable  of operation  by one man  (maximum weight
             40 Ib).
         2.   Self-contained.
         3.   Operable in a  safe  manner.
         4.   Operable in a  manner  such  that  no  permanent secondary
             environmental  damage  results.
         5.   Capable of  fabrication  from commercially  available  components
             to the maximum extent possible.
         6.   Capable of  reducing soil  penetration by 95% on  a land  area
             1200  ft? under 1  ft of  liquid  head for a  24-hr  period.
         7.   Capable of  sealing  many types  of soils under  varied terrain
             conditions  with minimum site preparation.
         8.   Capable of  application  in  varied weather  conditions.
         9.   Capable of  sealing  when applied before or  after spill  front
             has  passed  an  area; i.e.,  "dry" or "wet"  areas.
        10.   Capable of  preventing the spilled material  from seeping under
             the  edoe of the sealed  area.
        11.   Capable of  being removed  after the spill  incident is
             terminated.

    The program was designed (1)  to select and evaluate filin-producinq
sealing materials  that  could be used  to form a leak-free barrier on solids
in hazardous spill situations, and (2)  to design and fabricate a prototype
device to dispense the  materials  and Demonstrate its usefulness for
production.   The project completed only two of four scheduled phases when
it became apparent that design goals could not realistically be achieved.
No sealant candidate was found that could be  applied to essentially
unprepared terrain with any reasonable assurance of an effective seal even
unaer  ideal  conditions.  Polyurethane foam was considered the best of the
candidates surveyed.  But when it was applied to grassy areas, -it could  not
form a dependable seal  even at rates approaching 6.13 Ib/ft^—far in
excess of the O.r? Ib/ft^ rate of  the design  -joais, and impractical on
the basis of costs and logistics.

    Manufactured films that could  be used as  a barrier were  not  speci-
fically  excluded by the design criteria,  but  they were not to be  considered
in the primary program effort.   When & suitable film-producing sealant
could  not be found, however, our efforts  were directed towards other
methods  that might achieve  the program goals  in a practical  manner.

    Of the various commercial film materials  available, polyethylene  (PE)
was the  most economical  and had  the widest range of compatibility with the
representative list of hazardous materials.   Moreover,  it has widespread
commercial applications  and  is a common  item  of commerce.  The potential
was therefore extremely  good for developing  a spill-control  package
inexpensive enough for widespread  distribution  to the common carriers or
local  emergency response teams,  or for  ready  local  availability  to  a  spill
site.

    Polyethylene  film was first  investigated  as  a replacement for the
sealant  in the initial spill  scenario,  considering  its  suitability  for

-------
various terrain conditions, methods for deployment under adverse weather
conditions, and formation of suitable holding ponds from sheet material.
The concept proved to be practical.

    The capture-and-containment-bag concept evolved from the idea of a
polyethylene-film holding pond.  P> means was needed to capture the spilled
liquid and direct it to the temporary holding ponds, which may be some
distance from the spill.

    The original idea was to use a 1,5- to 3-ft-diameter plastic-film tube
with a capture apron on one end that could be pulled under and around the
spill, and to run the captured material dov.n the  tube to the containment
pond.  The containment-bag concept grew from a need to hold some of the
spilled liquid temporarily until the ponds could  be assembled.  To
accomplish this, the tube was enlarged for increased holding volume,
forming a  bag, and a small tube attached to the base of this bag for
eventually transferring the captured liquid to the pond.  The use of a
larger bag for both capture and temporary storage was a logical outgrowth.

    The work presented  in the body of  this report deals with the develop-
ment of the capture-and-containment-bag concept,  detailing  its development,
testing, and recommenaations for further development.   In our opinion,  it
is  a simple and  practical  first-response solution for controlling a large
percentage of  spills from  bulk  transport accidents.

    The initial  program effort  pertaining  to  sprayable  sealants  developed  a
significant body of  data that  were negative to  the  program  objective  but
Ore worthwhile reference information.   These  data are  presented  in  the
Appendix.

-------
                                  SECTION 2

                                 CONCLUSIONS
    The search for a soil-sealing system to be used immediately after an
accidental  transport spill showed that polyethylene film has the most
promise of bsing a basic barrier material.  Containment ponds formed from
?E sheet material and bags capable of holding 1000 gal were demonstrated
successfully under adverse soil conditions.  Polyethylene is readily
available, chemically resistant, and  inexpensive.  Moreover, its physical
properties are amenable to the  inclement weather conditions prevalent in
most transport spills.

    Bags formed  from 6-mil polyethylene film can hold more than a 1-meter
head of water without rupturing.  Reinforced polyethylene is available with
even greater rupture and  tear  strength.

    Heat-sealed  seams on  the units were the weak points.  Sewn seams,
though prone to  seep under pressure,  proved to be more dependable.  Double-
lined units with a heat-sealed inner  bag  and a sewn seam outer bag  appear
to be the most leak-free,  puncture-proof,  and strongest design for
containing spills under higher head conditions.

    A prototype  design consisted of a 20-ft-long, 6-ft-wide bag with  a
10-ft-long apron that collects the spill  liquid  and directs it  into the
bag.  Tie lines  on the apron are used to  position the device on the rup-
tured tanker.

    A transfer tube  attached to the bottom of the bag allows  liquid
transfer to a secondary container.  This  design  allows a modular  approach
to containing spill  liquid  in  several tandem devices.

    These bags can hold more than  1000 gal on reasonably  sloping  terrain.
They weigh less  than  16 Ib each, occupy less than 2 ft3,  and could  be
produced  in quantity for  $50 to $200  each, depending  on  the complexity.

    Field demonstration tests  of two  capture-and-containment  devices
confirmed tne modular approach. More than 1000  gal of liquid  were  col-
lected over 5 hr and stored without  loss.

    Polyethylene film  is  also  a suitable  barrier for  fabricating  emergency
holding  ponds.   Polyethylene film  6 mils  thick contained  water with more
than a foot of head  when  installed over weeds, reck,  and  other debris.
Larqe  sheets  would  be difficult to deploy under  wind  conditions.

-------
but mechanical  fasteners are available to join smaller sheets and thus form
barriers for larger pond areas.

    Use of film-producing chemicals for onsite sealing of soils wes not
practical.  Applying the film-forming formulations to substrates  likely  to
be found on berms and ditches near highways and railways produced barriers
that leaked through holes or seepage.

    Three film-producing classes were investigated:

         Notireactive sealants are  those  that  have  been previously
         polymerized ana are either dispersed or dissolved  in water or
         other solvents.  These  depend on  solvent  evaporation for film
         formation.  Thus they are slow  to  form at low temperatures,   if
         such nonreactive sealants are in  organic  solvents,  the vapors may
         pose a  fire hazard.  The  films  formed under  ideal  soil conditions
         often had  pinholes.   Leak-free  films were not possible over  rough
         terrain  and vegetation.
         Reactive sealants  are composed  of two or  more component  systems
         that react on  application to  form a  polymer.  They are more  likely
         to  form films  under  adverse  weather  conditions  than are  the
         nor.reactive sealants, but they  have  several  problems.
          Applications  to snow-,  ice-,  or water-covered  surfaces  cause
          frequent holes in  the film  and  result  in  poor  adhesion  to  the
          substrate. Applications  to vegetated  areas  often do not reach  the
          soil,  but form poorly-bridged films  with  frequent holes,
          especially adjacent to  the  vegetation.   Pinhole-size voids ere
          commonplace  in the films  formed even on  well-prepared  substrates.
          Surfactant chereic-.1 s consist of repellent crsf-^iicals that modify
          the surface  characteristics of materials  so that they are not
          wetted.  These failed to seal against candidate spill  liquids when
          applied to soils  under  ideal conditions.   Two of the irore
          promising surfactants investigated  (£.  I. duPont de Nenours'
          fluorocarbon  Zonyl RP and Clintwood Chemical Company's surfactant
          CUndrol 1,00 C& did not retain the  candidate spUl liquids when
          applied to well-prepa/ed soils at 22 to 25QC.  Even less success
          would  be expected on snow,  ice, rain-soaked soil, or vegetation.

-------
                                   SECTION 3
    The capture-and-contairr-,ent-t>i)Q concept as a f irst-resrinsr measure  for
the control of hazardous  spill  watt-rials trr-; bulk tappers should bf>
vigorously pursued.   The  units  are potentially inexpensive,  .irv.i t^Hr  use
in spill situations  is  reaciily  obvious to spill response r-rrsrnn.?) .  Tn.->y
are therefore attractive  candidates (or oni-oard twroency spi 1 1-ccnta tr,-;><'r)t
units fo.1 trucks and  rail  tankers or for distribution *nd i,se:  t>y  local
emergency response  teams.
    The basic  film  strengths of fiber-reinforced K fiHs  are r\ore
adequate  for the  intended use, Out sears nc^O aoditicna! study.   Sewn,  soams
a'~e strong, hut  they seep.  H^at-sealecJ rtco^> «r<»  leak-free, !:-ut  somet iw,
they  tail even at  low head pressures.  For tm-se reasons,  a cootie-walled
bag with  a  heat-sealed inner baq of cica--, '.ow-density  It  cortsir-ed  witn a
sewn-s^am outer  bag of f ioer-reinforced >JL ctfers  tne t-est promts*;  of  a
reliable  final design.

    Before  ootentirtl baq ^anuf acturori will  direct so.-irus efforts  to
solving  the several desicn prof>l<*ws «>vi«;f>nr». \<,  nwded  of  a p;n««nlial
mar<:>t.   Ihe  desian of t't? unit', snouM :••• evaluated  'n Hn«'  with
present-day baa  rt.anufacturinq technir,M<»", .  Duality control was  less than
adeciuate  on sea^ns and di^^nsions of the handmade test  units provioed by bag
suppliers.

    An expanded  proqram thould  include:

     (1)   improvements in deV1 li/ation,  etc.  This
          information can  then be considered  in  design  discussions wltn
          cofnr»ercial film  and bao  fabricators.
     (2)   Evaluation of cc^neroally ranut actured bags  to test the
          capability and rol i
-------
    Experienced spill-control personnel have recommended further evaluation
and testing of the concept.  Personnel from the Pennsylvania Department of
Environmental Resources in charge of spill control ir, the Pittsburgh
district observed a test of the fabricated bag.  They were impressed with
the simplicity of the concept, enough so to recommend in writing its
further development and to request units for testing on actual spills.

-------
                                  SECTION 4

                 MODULAR CAPTURE-AND-CONTAINKENT-BAG CONCEPT

CONCEPT DEVELOPMENT

    The capture-and-containment-bag concept evolved from a need  to capture
spilled liquid and direct it to temporary holding ponds constructed of
polyethylene film possibly some distance from  the spill.  The  original  idea
was to use ? narrow plastic-film tube with a capture apron on  one end that
could be pulled under and around the spill source,  and direct  the caotured
material down the tube to the containment pond.  The ban concept grew from
a need to hold some of the spilled  liquid temporarily until  the  containment
ponds could be assembled.  To accomplish this,  the  tube was  enlaraed  to
increase holding, volume, forming a  bag, and  a  narrow tube was  attached  to
the base of this bag to direct  the  captured  liquid  to the containment
pond.  The tube would be tied off  until the  pond was ready.

    With the success of the  initial  bag concept, the  idea of having
additional bags to  held captured  spill, rather than construct
polyethylene-lined  holding  ponds,  was  even  more attractive.   The use  of
bags  rather  than  the open  ponds gave better  vapor  control.   In addition,
bags  were easier  to deploy  than plastic  sheeting and needed  no filrc join-
ing,  hardware, or  diking materials to help  contain  the  liquid.  Th.?
assumptions  used  as a  design basis for the  final device were as follows:

    o   The total  available spill  material  may be as  much  as 10,000 aal;
         however,  in better than 45% of  the cases,  the  spill rate will  be
         such  that the  total spill  will  be .ess than 1000 gal.
    o   Containment  for  at least  1000 gal  will be made available in the
         capture-and-containment-bag for  quick, easy deployment by one
         man.  Subsequent  containment units will be developed in 600- to
          1000-gal  modules.
    o    For safety,  the  connection to the spill source will be made only
         once.   The capture system will  thus act as an emergency containment
          system,  but  have  the added capability  of directing any overflow,
          if  this  is required, to subsequent containment devices downstream.
    o    Diking  material  will not be available  for the deployment of the
          initial  first-action capture-and-containment system.   It might,
          however,  be  available for the development of subsequent downstream

-------
         containment areas, wiich could t.iKe the form cf additional baqs c»
         plastic-film-1Hed ponds.
    0    Procedures and/or a handbook describing techniques for connecting
         simple spill containment modules, or tiie use of polyethylene
         sheeting and sealant/fasteners tc forr. holding ponds w,!l t-e
         developed for in-field use by spill containment personnel.

    The final concept — the basics of which are shown  in figures  1 and 2
— consists of a polyethylene-1ined, piHod-shaped bag, about 6 ft wide and
20 ft long, with an apron on one  end.  The apron has rope lines attached to
each corner which allow it to  be  pulled through an existing spill  on the
ground, directing the liquid into the attached pillow bag.  Subsequently,
these lines would be used to pull tne ipron  into position under the
ruptured tanker and tied to the tanker in a  slightly elevated fasnior.  so as
to direct the hazardous liquid  leaking from  the tanker  into the bag.   Lines
located around the opening into the bag could be used for additional
support.

    The overflow 1s a 4-in -diameter polyethylene tube, approximately  30 ft
long.   It can be tied off near the bag, or  laid out  upstream  along the oag
to prevent  loss of liquid.  The tube can  be  used for emptyino the  bag  for
cleanup, or for directing overflow to additional storage downstream.

PROTOTYPE TESTING

    Most of  the problems with  the prototypes involved  the heat  seals,
especially  at the high stress  points where  the  transfer tube  was
connected.  Some of  this probably was due to the priorities  placed on  the
fabrication of these bags  by tne  suppliers.   Since  it  WAS a  dfvelopw?nt
-------
                                                       • I
Figure 1. Capture-and-containment device, showing dual-layer polyethylene
          film construction and collecting apron.
 Tigure  2.  Deployment of  device as used in simulated  field  tost.
                                     10

-------
    Low Pressure
    Regulator
Laboratory
Air  -
                                         Adhesive
                                             Tape
                           Low  Pressure
                       Gauge  (0-20  in.H20)
                                              Pressure
                                              Gauge
                                              (0-20 ps
                                                                 7  Heat Seal
 Copper  Wire
'(under  tension)
                                                                    Pressure  Relief  Valve
      1X]
                    Figure 3.  Apparatus used Co determine the strength
                           and integrity of polyethylene bags

-------
The results and comments are summarized  in  Table  1.

    The results indicated that  polyethylene bags  are  indeed  strong  enough
for the proposed use, but that  the  following  generalizations should be
considered in the fabrication of  any  capture-and-contairiment device:

    1.   Quality control procedures generally used for  ma<=s  production
         units would  be  insufficient  for tha proposed plastic film
         fabrication.
    2.   Heat seals  are  as  strong as  the base material  when  properly made.
    3.   Improperly  made heat  seals,  which  may pass present  quality control
         checks, may have devects which  would not hold  any liquid.
    4.   Plastic deformation  of 4 mi 1 (0.004 in.) thic* film occurs dur:r.«j
         the  first  3 hr  of  exposure to pressure heads of approximately 30
         in.  H20.
    5.   Failure of  these  same films required between 75 and BO in. Hj>Q
         for  rapid  failure.

Testing of Laboratory-Fabricated Clear PE  Bags

    Two prototype  pillow bags were fabricated from * mil  (0.004 in.)
polyethylene  plastic.tubing.   Standard 6-ft wide  (laid flat) tubing was cut
from  a stock  roll  ana one end heat-sealed  shut with  a hand-held
clamshell-type heat sealer.  The other open end  was  cut,  removing  half of
the tube for  about 2 ft to provide an apron.

     In the first test,  the bag was deployed  as a rolled-up  tube,
4 in. in diameter, at the  spill  site.   Water,  the simulated hazardous
 liquid, was fed from a  hose at approximately  1.5 gpm to  the bag.   As the
 water head built up in  the end of  the bag, the bag unrolled down.a hill-
 side with about a 10% grade  (1 ft/10 ft).   Full  deployment  occurred  in  less
 than a minute.  Water was  allowed  to flow  into the bag for  3 hr.  45  min.
 (340 gal) building  a maximum head  of 13 in.  with an  estimated  average  head
 of 8 in.  No chanrje occurred over  24 hr.

      In the second test, a  similar  bag,  but with  a 5-ft-long apron, was
 used.  The bag was  deployed  at a 45  degree angle to the downhill  slope of
 the  terrain.  This  position  was  chosen  *.o  determine the tendency for the
 pillow-shaped bag to roll  as  it  filled  if  not placed exactly downhill  on a
 sloping terrain.

      Again, the bag  was  placed  as a roll, with the apron deployed and
 attached  as  if collecting  liquid from under a ruptured tanker.  The mouth
 of the bag was held open  by  separating the film with several small rocks
 and  water was  supplied  onto  the raised apron at the rate of 5 gpm.  At the
 5  gpm feed rate,  water  first overflowed the bag without unrolling it, but
 with a slight  push  it  quickly unrolled with only a  2 in. head.   In b min,
 with 25 gal  of water,  the bag rolled sideways slightly.  After 22 min and
  110  gal of water,  further movement  sideways allowed water  to overflow the
  top  rather  than  run to the bottom of the  bag.   The  tie-downs on  the


                                        12

-------
TABLE 1.  RESULTS OF STRENGTH TESTS WITH COMMERCIAL POLYETHYLENE BAGS
                   16 in. x 9 in., nominal 0.004 in. thick)
Sample
1
2
3
4
5
6
7
8
9
Pressure
(in. H20)
15
30
30
110
35
75
30
30
80
30
80
30
30
30
85
Exposure
Time (hr)
1
3
0.5
< 0.0001
1
0.003
CO. 01
0.5
0.005
>0.01
0.0001
2.5
3
30
> 0.01
Mode of Failure
Did not fail, some deformation.
Did not fail, no further
deformation.
Did not fail, some deformation.
Failed through heat seal.
Did not fail, some deformation.
Failed through sidewall.
Poor heat seal.
Did not fail, slight, deformation.
Failed through sidewall.
Partial separation in heat seal.
Failed through sidewall.
Partial separation in heat
seal from deformation.
Failed through heat seal.
Did not fail; laboratory heat seal.
Failed through sidewall.
                                    13

-------
apron held the bag in place at that point.  Cutting the tie-downs from the
projected ruptured tanker allowed the bag to roll down the hill sideways
and empty in the process.

    A third test used the same bag, but with the bag deployed  straight down
the 10% grade slope.  The apron was held  16 in, above ground  level as  if
tied on the opposite side of the tank rupture, and water was  added at  the
same rate (5 gpin) to the apron.  It was again necessary to prop open the
bag mouth with a large rock so that the water readily flowed  between the
plastic film layers.

    In the first test, 315 gal of water were contained at a maximum head of
14.5 in. over a 20-hr perioo.  A total of 600 gal of water with a maximum
head of 24 in. was collected in the bag before failure occurred along  a
fold line in the polyethylene tube.   The  leak was approximately 6 ft from
the sealed end of the bag at a point  with about 20  in. of head.  Fold  lines
of polyethylene sheeting are notoriously  weak points.

Testing of Fiber-Reinforced PE Bags

    Bags fabricated  from fiber-reinforced plastic films were  evaluated to
obtain  increased strength and puncture  resistance.   Reinforced films  are
made up of a woven  plastic, mat-like  reinfjrcement  fabric  coated  on  both
sides with a thin film.  This construction  gives  the material high  tear,
puncture, and  abrasion  resistance,  the  degree  of  which  are  a  function  of
the density of  the  fiber weave.

    The first  unit  was  made  of  a  film fabricated  wHh 0.125  in.  vinyl
fibers  on a 0.25  in.  weave.   The  bag  was  8  ft  wide  (laid  flat) with  a 10 ft
apron,  a  20 ft  bag  section  and  a  30-ft-long overflow tube.   This  unit,
fabricated by  A.  Mamaux and  Son,  Pittsburgh,  PA,  was the  first with  a
transfer  tube  to  direct any overflow  to subsequent  containment devices or
systems downstream.

     The bag was double-stitch sewn, with  the stitching covered witii
polyethylene  mastic tape to prevent leakage.  It weighed approximately
16 15  and was;  relatively easy for one man to deploy.

     In  the  first test,  the bag was deployed as a rolled-up tube at the 10%
grade  test  site.   The bag was unrolled hy the single operator in freezing
temperatures  in 6 in. of snow,  and the transfer tube was deployed back up
the slope alongside the bag.  Water flowed into the bag at 5  gpm.  At a
total  of  450  gal  and a maximum head of 9 in., a leak developed in the
device at the sharp inside corner of the bag where the transfer tube was
 attached.  The leak was in the film fabric and not  in the sewn seam,  but
failure was thought to be caused by the stresses resulting from the sharp
 angle  at the  attachment point.

     Moving the transfer tube to empty the contained liquid from the bag was
 successfully carried out, but with difficulty.  The transfer  tube's flat
 width of nearly a foot was obviously too large, making it heavy when  filled
 with liquid.

                                       14

-------
    The leak in the A. Mamaux unit was repaired, and the size of the
transfer tube was reduced to 9 in. {'laid flat).  A radius was also included
al the point of attachment to reduce stress at this point.

    The bag was again deployed on the same downhill slope and supplied with
water at a rate of 5 gpm onto the raised apron.  Water height measurements
were taken every hour, as well as color photographs of the progress.  After
3 hr, 900 gal of water filled the bag to a length of 14 ft and 22 in. deep
at maximum head.  No appreciable leaks were observed; however, some seepage
was observed at the sewn seams.

    A second test with the same bag was conducted to determine the maximum
head the bag would hold on the 10% slope.  The apron was tied above ground
level as if tied un the opposite side of a tank rupture, and water was
added at the same rete (5 gpm}.  A total of 1350 gal was collected in the
bag in 4.5 hr before  a seam  began to  leak  (Figure 4).  The leak  again was
at the point of attachment of the transfer tube to the bag (Figure 5).
Water filled the bag  to a 17-ft  length with a maximum head of 26 in.

    The water was removed from the bag via the  transfer  tube.   The transfer
time was about 9.5 min, at a rate of  about 142  gpm.

    Although not completely  successful,  these  tests  indicated the
feasibility of the capture-and-containment-b?g  concept.   Deployment was
easy, and the containment safe to a  reasonable  head  of  liquid.   It was
obvious, however, that more  careful  attention  was  needed at  seams and
attachment points.

Testing of Fiber-Keinforced  Bags  from Commercial  Fabricators  .

    Commercially-fabricated  capture-and-containment-bags of
fiber-reinforced K,  similar to  the  ones tested,  were  solicited from
several commercial hag manufacturers  with  the  hope that  their  expertise  in
this  field  could c*d  in  solving  our  seam problem.   Their comments were
solicited on  the overall  design,  including the shape,  type of  fabrication,
and  seal  at  the  transfer  tube.

    Only  three  companies  agreed  to  fabricate  test bags,  none enthusias-
tically.  These  bags  were received  over  a period  of nearly a year and were
tested  very  similarly to  the test previously described.   Water, as the  test
 liquid,  at  a rate  of  5 gpm,  was  added to the bags secured on a sloping
 terrain of  about 10%.  Periodic  measurements were taken of the maximum
height  of water contained versus the volume of water added,  and the mode of
 failure,  if  any,  was  evaluated.   Modifications were then agreed to in
 conjunction  with  the  supplier,  and  the bag was retested.  A summary of
 these tests is shown  in Table 2.

          Griffolyn Company  supplied four bags made of Griff-Weave 24, a
 film with 0.125 in.  high-density PE fibers on a 0.125 in. weave sandwiched
 between 2-mil clear low-density PE film.  The total thickness was 8 to 9
mil.   The material  is exceptionally tough.  A tear strength of  750

                                       15

-------
                       Reproduced  (rom
                       best available copy.

U— »— "*

yr
1 II If
ft t
.

^*
^^
         v«


Figure 4.   Capture-and-containment  device with sewn seams  being
            tested to maximum containment pressure nead.
                                                     I








                                                   V/////A

 Figure 5.  Rupture of the device  in  the  transfer tube attachment  area.
                                       16

-------
                  TABLE  2,   SUMMARY Of  TESTS ON  FIBER-REINFORCED BAGS FROM COMMERCIAL  SAG MANUFACTURERS

1.
2.
3.
4.
5.
6.
Supplier
Griffolyn Company
Houston, TJC
•
•
•
SUco Industries
Newark. NJ
Katz Bag Coopany
Indianapolis, IN
Maximum Test
Type of Vtler Head
Fabrication (In.) Comments

Seams Manually
double-bar sealed
Double stitched sewn
seams - slllcone coated
Double stitched sewn
seams - PE-tape coated
Hot-air blown heat-sealed
seam
teat-sealed seams
Proprietary extrusion
process

12
12
29.5
<5
<5
48

Failed near the seam - heat distortion
Small leaks at jointed corners
Filled to capacity (~1SOO gal)
Seams leaked at about 0.4 gpra
Seams failed .zipper-like
Seams failed zipper-like
Bag on 1/4-scale
No failure
Weight
(IbL
13.5
--
~
12.6
9.8
«
   (full scale bag)
8.
Mechanical seal around
transfer tu'.e

Improved 0-rlng seal
around transfer tube
12       Leakage around mechanical seal


         Held for demonstration testing

-------
was reported for the film by the supplier.  In our own test, a slit that
was cut in a sample of the film could not be propagated with bare hands;
nor could a six-penny nail readily puncture it with only hand pressure.

    The manually heat-sealed test bag (No. 1 - Table 2) failed quickly at
about a foot of head, apparently as a result of heat-weakened film adjacent
to the seam.  Bags with sewn seams fared better.  One coated with silicone
(No. 2 - Table 2) leaked readily at the joints.  The one with the
PE-tape-covered searns (No. 3 - Table 2), on the other hand, did rot fail
catastrophically but seeped slowly at all seams.  Unfortunately, a number
of excess seams appeared to be present in the construction.  This bag was,
however, filled to capacity at about 1500 gal with a maximum head of about
29.5 in. (Figure 6).

    A hot-air blown heat-sealed bag (No. 4 - Table 2) was apparently not
well-sealed and opened up along the seams like a zipper early in the test.
The seal at the seam was apparently not hot enough to involve the high
density PE fibers, and the film just pulled apart.  A bag submitted by
Silco Industries (No. 5 - Table 2} with similar materials and fabrication
techniques also failed in the same manner.

    Katz Bag Company of Indianapolis, IN, a major fabricator of
polyethylene film bags, supplied a single-ply fiber-reinforced polyethylene
film fabricated on approximately 1/4-scale using their proprietary
extrusion process.  The bag was fabricated of polyethylene sheeting
consisting of t*o layers of polyethylene  (-vl mil) on either side of a
polyethylene fiber matt.  The sheet had a nominal thickness of 8 mil.
Other thicknesses are available.

    The bag (No. 6 - Table 2) was first tested to 2 ft and then to 4 ft
head.  Small holes appeared in the seams but none developed  into a
catastrophic failure.  The strength of the bag material was more than
adequate for liquid containment.  The material was abraided by stones  and
vegetation, end "pinholes" developed within the sheet  Hself during the
tests.  Neither the heat-sealed seams nor these abrasive cuts in the sheet
opened further, however, when the pressure head was increased to 4 ft.

    The full-scale unit (No. 7 - Table ?) supplied by Katz Bag Company used
a mechanical seal to attach the transfer  tube to the bag.  This seal
consisted of drawing the two pieces of polyethylene (bag and delivery  tube)
around a 1-ft  length of 3-in. diameter pipe.  PVC pipe was used for tne
prototype device, although polyethylene or polypropylene would be used in
the final design for better chemical resistance.  This device was tested
using the standard screening spill procedure and held only a 12-in. head of
water before some seepage occurred around the mechanical seal.
                                      18

-------
Figure 6.  Capture-jncl-contninnent device MJrtcated of
              •Griff-Heave 24".

-------
    A  tnird  unit  fr««,i Katz  tao tc'-'^any (Na. B - T.»;>lo _') incorporated  -,n
"G-ring" radif ication to their nnuinal  vpchanical sea) at Re trarsft-r
tube.   This  ba^  KJS  tested  success) u • ly during toe field uencnstrat io»i
using  a railroaa  tank-car simulated spill.
Jest ing of  iX j!)Jf ^j»j lie aq and 3 se»n reinterceu-r i:,er oi;ter aag construction
were solicits  Iron) co.'iwercia1  manufacturers.  Tne inner DJQ fomeo the
leJK-free  liner — tne outer bag provioed support for the higher
pressures.  Units  froT two companies »ere yurctusea.  Une fro*i Sheldshl
Corporation consisted  of  an inner bag of nonina! 6-rcil-th>CK polyethylene
film and an outer  bs<}  constructed of "?oly-Iut f ", a reinforced polyethylene
film.
             was  done  on  the sloping &rf>a usou for all screening tests.
Water was  directed  into  the ban it a QP»« with no signs of leakaqe until the
pressure heac!  was  increased to 31.5 in.  At that tine the plastic in tht>
area of the  seal between the containwnt baq and the delivery tube appeared
to lie strained,  ana tiv in. head for approximately 4i> mag In the section  between the delivery tu.be and the rontAinwent big was
o;>ened over  4  length  of  about ? ft, typical of poorly heat-sealed seaws in
polyethylene.  Although  the inner bag haa failed, the outer bag held well
under the  pressure.

    The project  was terminated before corrections and improvements could be
made in the  Sheldah)  units.  Ine prototype was of hich quality and it is
linely that  shortcomings could be corrected, tt-ne permitting.

    Stlco  Industries  of  Howark, *J. fabricated a 1/4-scale, double-walled
unit with  a  4  mil  ft  Inner "liner with a continuous heat-sealed seam, and a
Pt mesh outer  b^g  fabricated by a double-stitch-sewn-seam technique (Figure
8).  The guarter-scale unit tested successfully !o 46.5 in. of head with no
leaks (Figure  9).

    A full-scale double-walled unit submitted by Silco had »
double-st itch-sewn  outer bag of f iber-reinforcej HE film.  AUhouoh the
outer bag  was  to have been fabricated slightly smaller than thf- Inner bag
to give  immediate  support on filling, sjch was not the case, and the outer
bag provided only  a ioose fit.  On filling, the inner bag seam failed

                                       ^o

-------
Figure 7.  Capture-and-contalnment device fabricated
              by Sheldahl Corporation.
                                     21

-------
 i
Figure 8.   Construction details of  a  quarter-scale test
               device fabricated by  Silco  Industries.
Figure 9.
                                             Reproduced I'oni    | • J
                                             besl available topy.  V,,,j/
Quarter-scale  capture-and-containment device
      fabricated  by Silco Industries.
                                      22

-------
near the transfer  tube connection  after only  hit)  gal.

    Despite the above failure,  a unit  from  Silco  Industries  witn  this  scs^e
desiqn, but with close fitting  boqs-and n>ore  careful  attention  to inner  uao
construction, was  tested  successfully  in our  field  cieronstraticn.

>uTriarj^qf_^a^i|rj^jm1-Contairrwnt-Bag Concept  T>st ing

    Capture-and-containr'sent-bags appear to  be extremely  attractive  "first-
action" devices for  the control of eiwrgency  spUls of  lic,jios.   *jnit
capable of containing about  1000 gal wciqh  less than  ^0  lo and  occupy  a
storage volume of  less than  2 ff.3, depending  on the design.
Manufacturing costs  are estinateo  at less than  J20J,  with t> range.  In  ariJition, their deployment  is not
only simple, but obvious.  They are therefore attractive candidates for
onboard emergency  spill-containment units fcr truck ana  rail  tankers or  for
distribution and use by  local ome.-gency spill t;ams.

      Demonstration
    A field demonstration was  conducted with  two  prototype  capture-
and-containment-haqs  at  a site at MSA  Research Corporation,   The  simulated
accident was  a railroad  tanker Uid on its  side with  leakage  around  the  top
hatch (Figure 10).  Water was  the simulated spill  liquid  at a leak rate  of
5 gpm.

    The spill site  had only  a  qentle slope  away from  the  tanker,  which
limited the allowable head buildup  in  the baqs and thus total contuinnvnt
capacity.   It did,  however,  demonstrate the ease  of placement, thp utility
of the bags,  and the  ability to contain significant liquid  under  even these
conditions by use of  the modular bag concept.

    The fleTd demonstration  tested  two units  selected *s  th«  most prorr.ising
-- one fabricated by  S1lco  Industries, inc. and one from  Katz Bag Company.

    The SiUo Industries' unit consisted of an Inner  baq  fabricated  from 6
mil clear polyethylene film  by continuous heat sealing of all seams, and an
outer bag fabricated  from fiber-reinforced  PE film with double-sewn  seams,
The outer bag material was a close weave  1/8  in.  count *  1/8  in.  wide
polyethylene  fiber-reinforced  9-mH-thick polyethylene film.   Both bags  had
essentially the same  dimensions so  that the outer bag would retain the more
elastic inner unit  on filling.

    Grommets  for tie  lines were Installed at  several  positions along the
opening to the bag  (Figure 11) and  around the apron to assist in  deployment
and capture of the  liquid.   The lower  end of  the  bag, rather  than being
                                      23

-------
                               Reproduced liom    f •  )
                               t-.t-ii available copy. \_-_J
Figure  10.  Simulated tanker accident  for field  test demonstration.
                                        24

-------
Figure 11.  Detail of groinmets  for  attaching  apron.
Figure 12.  Detail of the  transfer  tube  attachment.
                                     25

-------

Figure 13.  Weight, size,and total volume of these devices
                   are easily handled by one man.

-------
cut and sealed square, was angled at about 45° where the transfer tube
was attached (Figure 12).  The delivered weight was about 15 Ib and
packaged volume less than 2 ft^.  It was easily carried by one man
{Fkjtrre 13).

    The Katz Gag unit was fabricated from single-ply 1/8 in. count x 1/8
in. wide polyethylene fiber-reinforced polyethylene film with a total
thickness of long delivery tube was mechanically attached
to the containment bag with an "0-ring" connection as  an alternative to
heat sealing.

    Effort was made to simulate a spill-response working atmosphere during
the demonstration to gain an  insight into problems of  deployment.  Safety
equipment was worn by the technicians, inducting respirators, safety
glasses, hard hats, gloves, steel-toe safety boots, and chemically-
resistant clothing.  To minimize exposure,  the  spill source  was approached
only once.by the technicians  to attach the  capture apron of  the  initial  bag
to the leaking tanker.

    The technicians approached  the  tanker,  with the simulated leak of  5  gpm
in progress, dragged the  apron of the  Silco unit under the  leaking hatch
(Figure 14), and tied  it  into position (Figures 15 and 16).  The  delivery
tube at the  bottom of the unit was  deployed uphill  in  order  to  serve  as  a
valve.  Figure  17  shows  the unit  in  final position with approximately  200
gal of contained water.

    After  positioning  the first  unit,  the Katz  bag was placed  in  tandem
below  the  first.   This  second unit  was kept in  reserve while the  first was
being  filled.   Because  of the sloping  conditions of the spill  site,  the
first  bag  was only able  to  contain  a  little more  than  500  gal  of  simulated
spill.  Figure  18  shows  the  liquid  level  in the bag at this  point.

    The  field  test was  continued  by transferring the  accumulated spill
liquid from the  first  unit  into the second  to verify  the Popular approach
to collecting hazardous  liquid from an accidental  spill.  The  transfer tube
of the first unit  was maneuvered without loss of liquid to the apron of the
second (Figure  19).

    The  sp1H  test was continued with the simulated flow from the tanker
 Increased to a  rate of 15 gpm.  Figure 20 shows the overall  view of the
spill  site and  bag deployment with  about 850 gal  captured.   The transfer
tube  of  the second unit was deployed upgrade to prevent liquid loss.  The
 test  was  terminated when the two units combined contained more than 1000
gal.   Figure 21  shows the condition of Bag No.  2 at this point.  No leaks
were  evident 1n either bag.
                                       27

-------
                               Reproduced  (com
                               besl available copy.
                   .
         „   _ -—^_  _ _ *>  ,K             J«— -  -» . ' •• J« -
Figure  14.   Field  deinonstrdtion test  -  deploying
              the capture-and-containment device.
Figure  15.  Field demonstration  test - securing the
                     device  to  the tanker.
                                      28

-------
                   •
Figure 16.  Field demonstration test - detail of the apron,
             secured to collect leakage around the hatch.
 Figure  17.   Field demonstration test - Silco Industries
                  device containing about 200 gal.
                                     29

-------


Figure 18.  Field demonstration test - Silco Industries
                 device containing about 500 gal.

       .
Figure  19.   Field  demonstration  test  -  modular  concept  demonstrated.
                                     3U

-------
                           Reproduced from   /"''"X
                           besl available  copy. \* 3

Figure 20.   Overview of the field  demonstration site  with  the
               two capture-and-containment devices deployed.
    "
                          •'•'"

 Figure 21.  Detail  of the Katz Bag Company device
                    containing about 500  gal.

-------
                                  APPENDIX A

                   INVESTIGATION OF  SURFACE SEALING SYSTEMS
GENERAL OUTLINE AND DESCRIPTION OF PROGRAM
    The initial program was designed to select and evaluate  sprayable
sealing materials which coula be employed  in hazardous material spill
situations and to design and fabricate a prototype application  system.  The
planned program consisted of the following  fc^  phases:

    1.   Laboratory Studies
    2.   Pilot-Scale Testing
    3.   Prototype Fabrication
    4.   Field Demonstration.

Laboratory Studies

    The first pnase of the program  consisted of  the  selection  and
evaluation of sealing materials which could fulfill  the  design  goals.   It
included a survey and selection of  sealant  candidates  and  chemical
compatibility studies and permeability tests that  defined  the  capabilities
and limitations of candidate sealing materials.

    A portion of the work at the beginning of  this phase also  was  dedicated
to the selection of representative  hazardous materials  used  in  the
compatibility tests.

Pilot-Scale  Testing

    The second phase of  the  program continued  the  evaluation of candidate
sealing materials under  conditions  that  simulated  those  of typical  spill
situations.  The sealing capabilities  of  the more  promising  candidates  were
measured when they were  placed  on  soil,  stone,  and grass substrates with
equipment  that could be  used  in accidental  spill situations.

Prototype  Fabrication

    The third  phase of  the program was  the design  and  fabrication  of a
prototype  device which  would  best  fulfill  the  objective of the program.
This  design  concept was  to be  based upon  the results obtained in the first
two phases and was  to be tested under  simulated spill  conditions to

                                      32

-------
determine areas fur modification and parameters  to  be  cmsidored  in  the
field dernonstr.it ', on.

Field IVynonstrat ion

    The fourth phase of the proorar.i was  to  be  a  field  test  of  the system on
a spill of IIJ.GGJ gal of simulated Hazardous  liquiu.

    Phases 1 and ? of tne program were conducted js planned.   At  the
conclusion of Phase ?, however,  it wa:- evident that sprayable  systems w?re
not the answer to the problem.   AS a result,  tht re"'aincer  of  the oriqit.al
progr.vn was aDcinooned, ana tne ervt'itas is  placed on the  c jpture-ano-
                concept describees  in tne main  body  of  tMs  report.
    This Appendix describes  the  spray  sealant  effort  ana presents the data
that constituted the major portion  of  1'iat  program.

DISCUSSION OF CANDIDATE  SEALANTS

    Recent advances  in  the development of  sealing 'Materials have
significantly increased  the  number  or  products commercially available.
This,  in conjunction uitn  the  broader  scope of applicability available in
these  new products, made it  mandatory  to survey the field for sealing
materials that could be  considered  canoidatei  for this program.

    The sealing materials  which  could  be considered for sealing soil
surfaces against  intrusion by  hazardous chemicals may oe grouped into three
general classes:  Nonreactive,  reactive, and surtac^-inochrying c'ip'-iicals.
These  classifications  indicate  the  chemical formation of the s«.>lant film
and/or the  interaction  of  the  sealant  with the substrate surface.

Nonreactive  Sealants

         Nonreactive sealants  are those which  have been previously
polymerized  and are  dispersed  cr dissolved In  either an aqueous or another
solvent system.   Such  materials  are primarily  thermoplastic in nature, and
Include such materials  as:

     1.  BUumastic
     2.  Rubber
     3.  Acrylic
     4.  Cellulosic
     5.  Fluoroplastlc
     6.  Phenolic
     7.  Polyester
     8.  Polystyrene
     9.  Polyvinyl  chloride
                                       33

-------
    Several (if  Uie.se "%Urt»iieh
t''t» oCo''Ouj r-etlia  rjKt'i i:0 (. . v/r\tr 1 :, .it iof   to  fif'jltfi  f ' :' SJTft/ '!.'.' fl.r;','. .  Ifif:,)1
^'jt'rus ois;.ers ions  are '.u:>i>-cv. to f rer/i'M:,  '•;!">*<• vt-r,  ar-.r: filr; t or-at ioi,
»'iic': depots upon evaporation, reflects  Uie  influence- ot ;.oth tf-turraturp
ana air velocity.  Mouseriold lau'x i.-iirts  jr»?  typical  rxi-ples; tftf rate o?
*il"; »orr'stijn  is  very slow it lower temperatures.

    Solvent svste-s  UCPO.MJ en ovaporjticn  of  I no  solvent  for til"
formation.  Solvents or solvent systems  otf'er  than  water  often nave
offensive oa;irs  or are fla~»Mable.  Many  solvents  will  contrikute to the
potential fire  cr  health hazards, urson L.SCU  in  rc'-Donse to spills or
    Tht>  films  prepared from dnv solvent  Systf'ns  aro subject to pinholc>
forr.ation,  even  on  very well prepareo Substrates,   nfien djipHcd ovor rouqh
or very  porous Substrates, formation ot  •»  Continuous f i IP Ln?cows very
difficult  to  (KCOiT:plish.

^ (.' J c t < v e  So a i in t s

    Reactive  sealants are two or more component  svsters in which one
component  is  reacted with a second component  (often with 4 catalv.t) to
yield a  polymer, 'fxanplcs or this «.ype  of  sealtng (naterial

          1.   fpoxy
          L*.   Unsaturated polyester
          3.   Phenol/ .'
          4.   t'roa/tomaldehyde
          b.
          Solvents are optional  In the  reactive  materials listed -itove.
Although  t^e  substrate r.$y be at a  low*?r  temper. iture, the Cfynponents of
these systems have 4 mtnifliu-ti temperature  for  nlxinq the components an
-------
                    A-l.   TOTAL  SEALANT  WEIGHT AS A FUNCTION  OF
                         APPLICATION HATE AND SEALANT THICKNESS
Application   Weight of Secant               Coating Thickness (in.) at
   Kite            Required                       Coating Density of
  (Jo/ft2)     (lb/'?uO ft*}                  (3 Ib/ft3)'(20  Ib/ft3) (60 Ib/ft3)
  0.03              36                         0.12         0.02         0.006

  0.06              Tl                         0.24         0.04         0.012

  0.10             120                         0.40         0.06         0.020
    Although foam expansion to 3 Ib/ft3 does not appear to be practical at
Viese low application rates, a sealant density of 20  Ib/ft3 will generate a
njcn thicker film with a greater possibility for sealing  than the unfoamed
materials at approximately 00 Ib/ft3.  The latter results in only a 6-mil
film at an application rate of -1.03  Ib/ft3 (the ncrmal thickness for
polyethylene film) and about a 20-mi 1-thick film for  an application rate of
0.10 Ib/ft3.

          Chemicals
    The third sealant typo consists of repellent chemicals which, when applied
to surfaces, modify the surface characteristics such  that the  surfaces are  not
penetrated.

    Extensive repellent technology has been developed for four classes of
materials:  textiles, paper,  leather, and masonry.   In each, there  is a  broad
range of techniques and chemical systems but only two classes  of chemicals  --
stlfcones and f luorocarbons •-  appeared promising for the proposed  use.  These
are used to provide the so-called durable finishes on textiles,  those that  are
essentially unaffected by either laundering or dry cleaning.

    SI If cone systems are employed either alone or in  combination with melamine
or ur
-------
againrt selected hazardous chemicals, and film cure at various
temperatures.  The application equipment requirements for each were also
investigated.

    The use of sprayable, film-producing chemicals for the onsite sealing
of soils was not practical.  Applying the film-forming systems to
substrates likely to be found on berms and ditches near highways and
railways produced films that leaked or allowed seepage underneath.

    A summary of the results for the three classifications is as follows:

    (1)  Nonreactive materials, those which have been previously
         polymerized and are dispersed or dissolved  in either an aqueous or
         other solvent, had the following faults:

         (a)  They are often subject to freezing and deactivation during
              storage or when applied.
         (b)  Film forming at Inwer  ambient temperatures can be very slow.
         (c)  Film forming at higher ambient  temperature can be too fast.
         (d)  When formulated with organic solvents, these systems require
              evaporation, which poses a fire or health hazard.
         (e)  Film formation often results in pinholes in the film as  the
              solvent evaporates.
         (f)  Leak-free film formation was not possible over rough terrain
              and vegetation.  A very uneven  film  with shadow areas having
              little or no film results.

    (2)  Reactive systems, composed  of two or more components that are
         reacted orisite to yield a polymer are more  likely to form films
         under adverse weather conditions, but have  several  problems:

         (a)  These systems cannot be used on snow,  ice,  and wet  surfaces;
              the spilled  liquid generally lifts the film and flows
              underneath.
         (b)  Holes in the film occur when the film  is formed on  top of
              snow, 1ce, or water.
         (c)  Applications to vegetated  areas produce  films  that  do not
              penetrate to the soil  below  and that leak  badly.
         (d)  Even on well-prepared  substrates,  pinhole-size  voids are
              commonplace  in the films.

    (3)  Repellent chemicals that  modify the  surface characteristics of
         materials such  that the surfaces  are not  penetrated  by liquids
         simply  were  not  effective on  soils.  Silicone  and  rluorocarbon
         repellent chemicals were  evaluated with similar  observations.

LABORATORY  INVESTIGATION

    The  laboratory  investigation of  potential  sealing materials was  taken
through  a  series of tests  designed to  outline their  capabilities and  the
possibility for  fulfilling the objective of  the  program.   The test  sequence
progressed  from  the subjective data  obtained  in  the  initial  survey to

                                      36

-------
pi lot-scale tests.

    This ;..riase of the procjra-i  "viluated  each  candidate  in  terms  of  the
conditions required to  form a  continuous  seal,  cnemical  ccTpatibi 1 ity of
the film with representative hazardous chemicals,  ana  o^rneani 1 ity  of the
film to liquid chemicals.   Fne  tests  *ere n.ide  under controlled  conditions
for comparison of trie various  typo-i of sealing  :-en sales representatives  to assure  an  Accurate appraisal of the
candiuate's performance  under adverse conditions.

    Samples were requested  for a  preliminary evaluation if, in the opinion
of us or the  supplier, fie  candidate  could neet the Jollowinq broad
criteria:

         A proi .?n  ability  to  fora a continuous film.
         Sprayabl« uider conditions practical for a portable  system, or
         could  be  wade so-wttn minor  modification to the fomulatlon.
    .    Able to form a  protective ikln  on the fi^m 1n a reasonable tlm*>
         period.   Fifteen  min at  made.-i.tf conditions appeared reasonable.
         Mo serious threat  Jn KseM  to  the operator or titc environment.

Manufacturers'  compatibility  data were essentially Ignored since most such
data are based  on  and  Intended for lonq-term exposures, whereas this
Intended use  was short-term.   Conditions such as bllsterinq or swelling,
which may affect a sealant's  use  In other applications, were  of little
hindrance to  Us use as  a  barrier to penetration of a chemical into the
substrate.

    Candidate sealants were screened by either soraylng or brushinq the
materials on  a  cardboard backinq.  Fhyslcal properties and set limes were
observed, and the  more promising  candidates were exposed  to a selection of
hazardous chemicals under  a foot  of head pressure.  This was  accomplished
by simply sealing  a foot-long section o* glass tube to the coated cardboard
with a  sillcone caulking as a sealant it the interface, and  Introducing the
test hazardous  chemical  into  the  tube.

                                       37

-------
           TABLE  A-2.   LISTING  OF  SEALANT  CANDIDATE  MANUFACTURERS
          Contact
  Potential Sealant
American Cyanaroid Company
Anderson Development Company
Ashland Chemical Company
Atlas Minerals and Chemicals
Borden Chemical Company
Gallery Chemical Company
Carboline Company
Ceilcote Company
Clintwood Chemical Company
Corrosioneering, Inc.
Dow Chemical Company
Or. Leonard Spialter
duPont Specialty Chemicals
E S B Incorporated
General Electric
Goodloe E. Moore
Hooker Chemical Company
Irathane Systems, Inc.
Johns-Manvllle
Koppers Company
Lastek Paint,  Inc.
Locktlte Corporation
Mameco, International
Midland Adhesive and Chemical Corporation
Onelda Electronics Corporation
Orb Industries, Inc.
Palmer Adhesive*
Urea/formaldehyde
Silicones
Urethanes
Adhes'ives
PVC copolymers
Urethanes
Vinyl paint
Bitumastic and polyester
Surfactants
Epoxies and polyesters
Epoxies
Silicones
Fluorosurfactants
Adhesives
Silicones
Adhesives
Phenol/formaldehyde
Urethanes
Ceramic adhesives
Polyesters
Bitumastic coatings
Adhesives
Urethanes
Epoxies
Adhesives
Aerosol adhesives
Adhesives
                                      38
                                                                  (continued)

-------
            TABLE A-2.  LISTING  OF  SEALANT  CANDIDATE  MANUFACTURtRS
                                 (continued)
          Contact
  Potential  Sealant
PPG Industries
Pocono Fabricators, Inc.
Pullman-Kellogg Company
Protex-a-Cote
P.V.O. International
Randustria1 Corporation
Royal Industries
Sauereisen Cements Company
Shell Chemical Company
Shiloh Specialty Company
Tracor, Inc.
3M Company

Trowelon,  Inc.
Union Carbide
Polyesters
Cements
Fluorocarbons
Coatings
Proprietary aqueous system
Urethane sealer
Fluorocarbons
Cements
Surfactants, epoxies
Sealants
Previous work
Polyesters, surfactants, and
 urethanes
Polyesters
Polyphenol and phenol/
 formaldehyde
                                      39

-------
    The oreliminary challenge chemicals were wator, sulfuric acid, sodium
hydroxide, trichloroethylene, methanol, naphtha, and methyl ethyl ketone.
Conclusions as to the utility of the specific sealants based on the results
of exposure to these challenge chemicals are summarized  in Table A-3.
Failure of the sealant coating generally was catastrophic with loss of
liguid.  For those cases where failure was observed, similar chemicals  in
the classification were tested for verification.   Those  considered for
further testing generally passed all preliminary tests.

    Although representatives of all  13 sealant  types were evaluated,
promising candidates were limited to only  four  of  these.  They were:

         Adhesives           Borden Chemical Company's Polyco 2607
         Urethanes           Gallery Chemical Company's  Resin 115
                             Ashland Chemical Company's  EP 65-86/88
         Surfactants         duPont deNemours1  Zciyl HP
                             Clintwood Chemical Company's CUndrol 100CG
         Plastic Sheeting

    In general,  it was  felt  that aqueous  sealants, repellent  chemicals, and
adhesives required additional experimental  testing to  provide more complete
data  for evaluation.  For the reactive sealant  candidates,  only  the
urethane systems were found  to combine mos4.  of  the characteristics required
of the sealing materials for this  program.   The epoxy  systems were found  to
have  excellent chemical compatibility  when exposed to  the challenge
liquids, but  required special care  in  sett'ng  the  component  ratios and
thick  layers  for suitable field  application.  polyester  systems  were,
generally, quite easily sprayed, but  ,iad  severe temperature  and  chemic.il
compatibility  limitations.   Un?  silicone  formulation was found  which  could
be considered  for  field applications,  but M^h  viscosity and limited
chemical compatibility  would restrict  its sealing  capabilities.  '

    Several  commercial  urethane  formulations were  found  to  be generally
applicable for  the  intended  use.   The  considered  opinion concerning  their
me was  that foaming formulations  would  provide the best combinution of
characteristics  to  fulfill  the  objective.

    Several  types  of plastic sneeting were found  to have general appli-
cability for the intended  use.   The initial attitude concerning their  use
 in  this  program,  however,  was  that they were not  of primary interest,  and
that  sprayed systems were  to be emphasized.  This attitude was later
revised  and  polyethylene sheeting in the form of bags became the main
thrust of  the program.

    At a later point in the program, since polyethylene film proved  to have
such  excellent compatibility against most challenge agents, polyethylene
 and  polypropylene  prepolyrners  were tested as a sprayable film.  These  were
obtained as  low melting waxes  that were melted and sprayed on a substrate
 to cool  and form a film.

-------
TABU A-J.  SUWt  OF  SEALING MATERIAIS
Set I ant Type
Adhesive


Acrylic
Paint
Silicon*
Urethane
Ur«thtn«
Surfactant
Pol/ester
Coivvty
Goodloe C. Hoore
One Ida tlectronlcs
Borden Chemical
Johns -**nvi Me
Emerson I turning
Cote-P.an»er
Ctrboline
General tl» 65-46/88
S«vle kit
2onyl Kf
Cllndrol IOOC6
Flrteline 75?
1070 Resin
General
Perfor»aiKe
Questionable
Quest (nntble
Questionable
Poor
Good
Good
Good
Question able
Quest (enable
Questionable
Questionable
Poor
Question aole
Questionable
Good
Questionable
Sood
Good
Questionable
Good
QuK.tlonable
bood
Poor
Good
Question able
Conclusions I Observations
Unacceptable compatibility
Unacceptable compatibility
Unacceptable compatibility
Unacceptable compatibility
Questionable
Unacceptable
Acceptable; slow degradation
Unacceptable; core conditions
Unacceptable compatibility
Unacceptable compatibility
Uniccepl*l>l£ compatibility
Unacceptable compatibility
Unacceptable compatibility
Unacceptable vtscos!*.- and
compatibility
Unacceptable viscosity
Unacceptable compatibility
Ursccepl'hle viscosity
Acceptable
P*ior fl'alnq
Acceptable
Unacceptable cowpat fbllity
Quest lonaMe
Unacceptab'e convatibilit/
Unacceptable viscosity
Unacceptable compatibility
                                                        (continued)

-------
                             TABLE A-3.   SURVfT  OF HALING MATERIALS (continued)
See lint Type
Aqueous (two
component)
Epoxy
Ure«/
formaldehyde
Phenol/for -alde-
hyde
Plastic Ft|«s
I Sheeting

Company
Pacific Anchor Otv.
P.».0. Internat't
Corrostoneerlng
American Cyanantd
Hooker Che* leal
Aero-Jec
«!lied Cheaical
Foaned Poly-
Product *JO*
Chew-Cure
RestsU-Flafcr 4200
—
~
Polyurethane
Polyethylene
Polypropylene
Foamed Poly-
ethylene
propylene
Fowd Poly-
styrene
General
Ferforwrxe
Questionable
Questionable
Questionable
Questionable
Good
Good
Good
Good
Good
Questionable
Conclusions t Observations
Questionable
Questionable; thick coatings
required
Lnacceptable; added pollution
Unacceptable; added pol"it)on
Acceptable
Acceptable
Acceptable
Unacceptable; hot preparation
required
Unacceptable; hot preparation
required
Unacceptable; compatibility
Cerent
Pocrno Fabricators

-------
Selection of Representative Hazardous Materials for Challenge Agents

    Recent growth in the use and transport of chemicals has significantly
complicated the selection of materials for representative challenge
agents.  The number of commercial chemicals and the degrees of hazard which
they may present to the public or the environment make this selection
difficult, at best.  Selecting the materials to be used for evaluating and
developing a system for sealing soil surfaces required a thorough search
and judicious analysis of data generated in many diverse areas.

    The selection of representative hazardous materials was made from data
published by chemical manufacturers, safety associations.,and both
government and insurance agencies.  These sources were used to select 30
chemicals to represent those that are transported in bulk and that would,
if spilled, present a significant threat to ground wf»ter supplies in the
area.  These materials are presented in Table A-4, which includes
representatives from most chemical classes.

    Data, such as the quantity transported, relative hazards such as TLV
(Theshold Limit Value), LC (Lethal Concentration), and LEL  (Lower Explosive
Limit), spill probabil Hies, and biodegradability were used  as criteria in
selecting these chemicals.  When overall ratings were similar for several
candidates, the final selection was based upon chemical aggressiveness
toward containment systems.

    A typical example may be shown for the  inorganic acids  class.  Sulfuric
acid is an obvious choice, based on the quantity which is transported.
Nitric and arsenic acids were possible second choices.  Although it was
recognized that arsenic acid represents a greater pollution hazard, nitric
acid wes chosen because of its greater chemical reactivity  toward most
sealing materials.  The logic in the selection was that any system capable
of containing nitric acid would  also contain arsenic acid.

    Sodium hydroxide was chosen  to represent all three major  inorganic
bases:  sodium, potassium, and ammonium hydroxides.  All are  aqueous solu-
tions posing essentially the same reactivity and thus containment problem.

    Phosphorus trichloride is a  typical liquid  inorganic halide
transported.  We have not  included any solids,  such as aluminum  chloride,
as challenge agents even considering that  in a  rain water-soluble solids
might be considered a hazard.  We believe  that  in a rain, where  solids
could be a problem, the obvious  solution  is  to  place a protective cover
over the spilled chemical.

    We chose phosphorus oxychloride  as our  inorganic oxidizer represen-
tative, although nitric acid also would be  classified  as  such.   Silicon
oxychloride was an alternative choice.  Others  are also  handled  in  bulk  --
namely hydrogen peroxide,  calcium hypochlorite,  and sodium  bichromate.

    Carbon disulfide was chosen  as  the organic  sulfur  derivative because  of
Us solvent properties.

                                      43

-------
          TABLE A-4.  PRIMARY REPRESENTATIVE  HAZARDOUS  CHEMICALS
Inorganic acid:
              Sulruric
              Nitric
Inorganic base:
              Sodium hydroxide
Inorganic halide:
              Phosphorus tri-
                chloride

Inorganic oxidizers:
              Phosphorus oxy-
                chloride
Organic acids:
              Acetic
              Acrylic
Organic halide:
Aldehyde:
              Methylene chloride
              Trichloroethylene
              Acetaldehyde
              Benzaldehyde
Plastic monomer:
              Styrene
              Vinyl  acetate
Alcohol
Ester:
Amine:
Ketone:
            Isopropanol
            Cresols
            Ethyl acetate
            Ethyl acrylate
            Ethanolamine
            Pyridine
            Acetone
            Methyl ethyl ketone
Mixed solvent:
            Naphtha
            Turpentine
            Lacquer solvent

Mixed hydrocarbons:
            Gasoline
            Kerosene
Aromatic:
            Toluene
            Xylene
                                     44

-------
    Two organic acids were selected — acetic arid acrylic.  Acetic acid has
the most spill potential.  Acrylic, on the other hand, has a  significant
potential for spills based on intra-plant transport and poses  a  tougher
containment problem.

    Methylene chloride and trichloroethylene, because of their excellent
solvent properties, were chosen to represent the organic halides.

    An organic sulfur representative was not included.  The mercaptans are
transported in bulk, but present no unusual solvent or reactivity
properties.

    Acetaldehyde, benzaldehyde, and acrolein were selected over  furfural,
butyraldehyde, and formaldehyde.  Each candidate has either high spill
potential or poses a difficult problem to the sealant.

    Isopropanol and a cresol were selected over etrylene glycol  and  ally!
alcohol.

    Herbicides and pesticides '-/ere not selected as  challenge  agents.  To
our knowledge, none are transported as bulk liquids and, when formulated  as
liquids, are either in aqueous or hydrocarbon solution.  Thus these
classifications are best represented by the solvent.

    Ethyl acetate and ethyl acrylate were chosen to represent esters.
Vinyl acetate and methyl methacrylate are possible  alternatives.

    Ethanolamine and pyridine were chosen over  acrylonitrile  and aniline  as
organic nitrogen derivatives.  All are good choices, but ethanolamine  and
pyridine present fewer problems in laboratory handling than acrylonitrile
and aniline.

    For ketones, acetone and methyl ethyl ketone were  selecte I.

    Mixed solvents pose a  significant hazard to most sealant  membranes.
Naphtha, turpentine, and lacquer solvent.were selected —  the latter
consisting of a mixture of  acetates and ketones.

    Gasoline  and kerosene  were chosen to represent  mixed hydrocarbons  as  a
class.  Other alternatives  include JP-4 and cyclohexane.

    Of the aromatics, toluene  and  xylene appear to  be  good
representatives.  Possible alternatives are phenol  and tetrahydro-
naphthalene.

    In a last category,  styrene and vinyl  acetate  were selected  as plastic
monomer candidates  that  have a high spill  potential and  pose  unique
reactivity problems.
                                      45

-------
Laboratory Test Procedure

    The sealing materials selected from the survey as potential candidates
were tested for their sprayability, film-forming capability, ana chenical
compatibility.  These tests were carried out  in conjunction witn the
manufacturer to test each sealant under- its optimum conditions r'or good
film formation using suitable equipment, conditions, and optirnun ratios of
components.

Sprayabi1ity Tests --
    Sprayabi1ity tests were conducted  in a small spray chamber, a schematic
of which is shown in Figure A-l.  The  unit consisted of a variable-speed
belt conveyor with a speed range of ?0 to 120 ft/min, a spray  gun with
supporting and controlling equipment,  and a vented spray cabinet for
controlling fumes and oversp^ay.  All  fumes were exhausted  through an
activated charcoal filter to capture organic  vapors.

    The spray gun varied with the sealant.  Single-component sealants were
applied using a single-component UeVilbiss Model 504 flat spray gun.
Two-component sealant candidates were  applied with a Binks  Model 18 FKA
gun, which also provides a flat  spray  pattern.  The Binks gun, which
includes an internal mixing chamber, is especially required for the
two-component urethane sealant candidates because of short  cure times.
Spray rates were kept nearly uniform throughout at about 1  to  3 Ib/min,  and
application rates held in the range of 0.02 to 0.1 Ib/ft^.   These rates
are considered in the probable range of the final piece of  equipment.

    The substrate was asbestos paper about 0.012 in. thick, which provided
a uniform surface for good film  formation and a porous substrate for
subsequent compatibility testing.  Six-in.-diameter sprayed samples were
prepared and  then mcunted in the gas pressurization and compatibility  test
units for evaluation.

    Initial attempts  to  spray  the  candidate were done with  flowmeters  in
the system to measure and monitor  all  the liquids.  After establishing  the
precision of  flows and stoichiometry,  the liquid flows were calibrated  by
weight  and assumed to be constant  throughout  the 3  to 5 min experimental
spraying times.

    Gas  pressurization was considered  the best means  for delivering  and
metering the  sealants.   Between  50 and 90 psig was  used  in  the tanks
containing  the sealants  to obtain  the  desired flow  rates.   In  addition,
about 1  to 2  cfm  air  was added  at  the  nozzle  to  supply  atomization  air  to
help set  the  spray pattern.  A  "flat  V"  pattern was  found  to produce
uniform application  rates on sample  surfaces.

Compatibility Tests  --
    Tests were conducted initially on  small,  provisional  samples  and,
later,  with  the chemical compatibility test  unit  shown  in  Figure  A-2.
                                      46

-------
                    EXHAUST
SUPPLY  GAS
  5 GAL. POTS
                                            EXHAUST  FAN
                                               FLAME ARRESTER
                                               4- FILTER
                                                      CONVEYOR TABLE
SPRAY  CABINET
                     Figure A-1.  Schematic cf spray chanber

-------
                          ir
       TEST FLUID
           BURETTE-
    -WING NUT

                BLEED
-TEFLON  \     PLUG-7
 -. BLOCKS    \	J
                              STCP  COCK

         PAN-

            Hguro A-2.  Compatibility lest unit
                                             LIQUID  LEVEL
                                             |N.     CM.
                                             12 — 30.4

                                             H  — 27.9
   9

   8

— 7


   6

  • 5
   3

   2
                                              0  —
25.4

22.9

20.3

17.8

15.2

IZ7

10.2

7.o

51

^5

0

-------
This apparatus, fabricated from Teflon^stock,  was  designed  to  expose  a
large surf ace area of sealant  (12.5  in?)  to  a  1  ft head  of  ch-ilenge
liquid, while using only a mininum volume of chemical.   The inside of the
top section wi* domed for qas  escape and  exposure  of  liquid to the total
sealant surface.  Tne natinq base section is an  open  cylinder, with a 4  in.
instdc   ml.

Film Forminq  Capability  —
    Tne  integrity of tne film  was  c.v'&luated  by measuring the  leak rate of
air U'rouqn the curt-d sample mounted in the  base of the compatibility test
unit prior  to the addition of  challenge aqent.  The apparatus, shown in
Figure A-.J, allowed  pressurization  tc a l?-in. water head.

Test Procedure  --
    Tne  U">t  procedure  was as  follows:   Conditions were set up and checked
{flow rates,  stcicniometry,  conveyor speed,  and cpray pattern) on  the spray
           to  achieve a  suitable  film for the sealant under study  at  the
         application  rate.  With  the asbestos substrate on  the base of the
         i IHy tost  unit  and  preconditioned to the desired  temperature, the
         wen?  placed on  thv movinq conveyor and the sealant r.pplitid.  They
w-.-ro  t'>"n  r'-turni'rt  *,o  the  des>qn temperature condition and the curing rate
monitored.

     The  t«»sts w^-re  conducted  in  multiples of 4 to 8  units.  This  number
allowed  tnrtividual  monitoring of cure  time, and also allowed  us  to assemble
the compatibility test  appa-atus arid to expose the film to the challenge
 liquid  in  a reasonable  tin*  period  (15 to 30 min).

     After asvembly of tne  compdtibUity  test  apparatus  and before the
 addition of challenge chemical,  the  integrity of  the film  and seal was
 tested  against g-js  pressure.  Generally,  acceptable  samples  leaked in the
ranqe  of 0 tc 2 in./">in of water,   Samples  that leaked  at  a rate greater
 than  4  In./mm were not subjected to the  challenge chemical.

     The  hazardous challenge chemicals  were  added  to  the  test  unit to a 12
 in.  head in a fune hood.  The lea»,  rate  was monitored  in the  burette
continuously during the first hour,  and  at  intervals throughout  the
 remainder cf the day.  Lost liquid  was replaced periooically  to  maintain a
 \2 in.  nead.

     The test period was scheduled to run overnight,  during which the
 samples  were unattended.  If  the samples exhibited a leak  rate within the

                                       49

-------
Plant A1r
               Low Pressure
                Regulator
0-30 in. water
                               Toggle
                               Valve
                                                          water
                                                          Manometer
                                               Test Specimen
             Figure A-3.  Air pressurization apparatus used to screen  test  specimens

-------
allowable limit (approximately f> nl/nr) they were left expired.  Those with
ledk rates in excels or b cil/nr hi<3 ttse liquid head removed t>y closing the
stopcock, and the head replaced in the morning to obtain additional
approxinate data.

    For comparison, the  !e:>k rate data ire reported as LpJL:3! L:2iii!i
         TOO for ,'4 hr (ml/?4 hr).  (his required 'extrapolation of short
term test uata to 24 hr for the many samples with high leak rites.

    All samples, whether failures or successes, were opened in the hood and
Checked for visible changes.  Holes were usually noticeable in test films
which failed,  tven saccesstul r :n<. often nad discoloration, Blisters, and
swell inq, indicative of some attac* L~y tht cher.ncal.

Uisciission of Test Results

Results with Adhesives —
    This general classification contained many chemical types, but most
were eliminated  in the preliminary survey due to d known  lac< of  chemical
compatibility.   Latex  adhesv/es were the exception and ream red  turtner
testing.  A Borden system, Polyco i'fcu/, was  found that formed a  film within
minutes and was  shewn  to he compatible with  nany organic  materials.
Samples of this  were tested on asbestv.  paper substrates.   The result* are
compiled in Tjble A-5.

Results with Repellent Chemicals  — •
    The preliminary survey of potential sealing materials  revealed that
some materials  in this classification may hav? the ability lo modify the
surface characteristics of the substrata and reduce  penetration  by
ha/ardous materials.   Toe survey  concluded  that two  types, floorinated
polyacrylates and superaoides, were potential candidates.

    Several commercial samples received some experimental  investigation,
but only two, E.  1. duPont's  fluoritiated polyacrylate  "Zonyl k?" and
Clintwood Chemical's superamide  "Clindrol  100CO," were subjected to  tests.
Neither showed  any significant sealing ability when  applied en  sand
supstrates at rates ostween 0.01  and O.Ob q/ci^, even  aqainst water.   On
asbestos, however, some sealing  ability was  demonstrated  for a  limited
time.  The results for Clindrol  are shown  in Table A-6.

    The liquid  transmission rates represent  3D to 80%  reductions over
untreated substrates for  the  limiteo  times  shown; however, the   samples
appeared to show a general  solubil'ty of the Clindrol  100CG in  the
challenging  liquids.   When  the costing was  dissolved,  transmission of  the
liquid occurrea at the same rate  as with untreated  paper.
     The  results  for  Zonyl  HP  are suwmarized 
-------
TABLE
RESULTS Of CCMPATIBlLIT*-?tRK£A8Il.lTY  TESTS *ITH S&RD£S
CHEMICAL CCKPM*  LATEX -Pa''CO  2607*
Application Condi
Ch«l tetiq* Liquid f
4«ou*.
Hftnylmo cnlono*
L^cgucr ialvent
Styr*««
Carbon QtlulfliM
PHciptorui 01 yen tor 1
Iyiff»«i
Cruol
Nttnyl «thyl ktten«
«.»»..
So«1u* lyoroxld* 15X
Afttitdenyd*
j^lfurK td*
Hono*t*t«no 1 M1M
Uoproo*no)
Vinyl Ktm*
^ftotonoruf trtcnlorl
MtttP
Vtnyl »Ctt*tt
•UAtlK
•Hpf.Wt
S*toHf»#
&O
-------
      TABLE  A-6.   TEST  RESULTS  OBTAINED FROM SAMPLES OF "CLlNDROL  100CG"
Challenge
Liquid
Benzaldehyde
Naphtha
Trichloroethylene
Average Liquid
Transmission Rate
(ml/min)
3 to 4
2 to 4
4 to 5
Breakthrough
(min)
5
20
5
Urethane Foam Systems Results —
    Callery Chemical's Resin  115 and Ashland Chemical's  EP  65-86/88 were
further tested as representative urethane  foam candidates.   Application
rates between 0.02 and 0.12 g/crr? were  tested on  asbestos after  it^was
established in preliminary tests that rates  in excess  of 0.02  g/cm2 were
necessary to seal the sand surface.  The compatibility test results are
suir-narized in Tables A-3 and  A-9.

Polyethylene Prepolymer Results —
    The chemical compatibility of polyethylene with  many hazardous
chemicals was considered a definite  attribute for its  u^e as a sealing
material.  A thorough search  of commercial  forms  of  polyethylene revealed
thet low-density polymers are available which could  be used to form a film
in situ.  These materials, often referred  to as  polyethylene waxes, have
the characteristic low softening points and  low  viscosities commonly
associated with natural waxes.  The  possibility  of spraying polyethylene,
with prior heating, was investigated as well.

    Three samples of these materials were  supplied by  Allied Chemical
Company.  Some selected properties of  them are  listed  in Table A-10.

    The visco-elastic flow properties  of  these materials make the softening
point and melt viscosity  important factors in the choice of materials for a
spray application.  Lower softening  points tend  to lower the system
operating terrperature, which  means that evaporation of all  system com-
ponents will be minimal.  Lower melt viscosities tend  to reduce the  power
requirements for spraying, which translates  into payload and cost savings.
Both of these are desired, but unfortunately the capacity  for forming
continuous films Is  inversely related  to  softening point,  viscosity,  and
density of polyethylene polymers.


                                      53

-------
TABLE A-7.  RESULTS OF COMPATIBILITY-PERMEABILITY TESTS WITH
            E. 1. DUPONT Oe NEMOURS SURFACTANT "ZCNYL RP"
Application Conditions
Approximate
Challenge Liquid Temperature, °C
Trichloroethylene
Trichloroethylene
Gasoline
Vinyl acetate
Acetic acid
Water
Water
Naphtha
Trichloroethylene
Acetic acid
Vinyl acetate
Naphtha
Water
Acetic acid
Ethyl acrylate
Monoethanolamine
Cresol
Gasoline
Methylene chloride
Benzaldehyde
22
22
22
22
22
22
22
22
25
25
25
25
25
25
22
22
22
22
22
22
Substrate Application Rate
Condition (g/cm2)
dry
dry
dry
wet
dry
dry
dry
dry
wet
wet
wet
wet
wet
wet
wet
wet
wet
wet
wet
wet
0.03
0.03
0.03
0.03
0.03
0.03
0.02
0.03
0.003
0.003
0.02
0.02
0.02
0.02
0.01
0.01
0.01
0.003
0.003
0.003
Total Liquid
Transmission
(ml/24 hr)
&
140
5
3
>!000
380
225
9
60
>1000
43
19
6
27S
25
63
13
7
7
28
                            54

-------
TABLE A-3.
RESULTS OF COHPATI8ILIT»-P!.RM£ASlLin TESTS WITH CALLER'
CHEMICAL COMPANY URETHAWE "1)5" FORMULATION
Appt tc»t1on Conditions
Aoorax1««te
Challenge Liquid Te*t>«rAtur»,oc
Aerolein
Carbon dlsu'flde
Pyrldlne
SodliM hydroxide I5N)
Gasoline
Phosphorus trichloride
Phctpnorut oxychlor^de
Cresol
Acetaldetiyde
Carbon d1su1f1<1e
Mono«ttianol«mlne
Sulfurlc acid
Trlchloroethylene
6«nz«ldehyoe
Acetone
M«thvlen« cMorlde
'thyl acrylate
Acetic *cid
Styrene
Hethylene cftlorlde
Isopropanol
Acetone
Cretol
Xylene
Gtso'dne
Sodium hydroxide (SM)
Benzaldenyde
Trlchloroethylen*
Sulfuric *c1d
Vinyl acetate
M.N *01metnylfonMMlde
22
22
22
22
:z
22
22
22
22
22
22
20
20
20
20
20
20
20
20
•5
•5
-5
•S
-5
-5
•2
•2
•3
-3
-3
-5
Appro*. Total Llould
Substrate Application Kate Cure Transmission
Condition (g/cnJ) Tlme(mln) (mV2< nr)
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
wet
net
»et
wet
wet
dry
wet
*ry
wet
dry
wet
wet
0.02
O.C2
0.02
0.02
O.OS
0.05
0.05
0.05
0.05
0.02
0.03
0.05
0.05
O.Cs
0.05
0-.12
O.U
0.12
0.03
0.04
0.04
0.05
0.05
0.04
O.OS
0.04
0.04
O.OS
0.07
O.OS
0.05
2
2
2
2
2
2
2
?
2
2
2
2
2
2
2
5
5
5
1
25
25
10
10
10
5
25
IS
20
10
20
20 fatUd
'.9.1
4.9
8.7
12.7
0
8.2
14.2
11.3
44.5
4.9
2.1
4.2
2.4
21.1
failed
23.0 '
6.3
7.7
2.1
7.6
0
30.0
2.3
1.6
S.9
3.6
S.I
4.6
14.5
20.9
after 2.5 h?

-------
       TABLE A-9.  RESULTS OF COMPATIBILITY-PERMEABILITY TESTS WITH ASHLAND
                    CHEMICAL  COMPANY  "EP65-86/88"

Appl
Approximate
Challenge Liquid Temperature, °C
Water
Sulfuric acid
Sodium hydroxide (5N)
Phosphorus trichloride
Phosphorus oxychloride
Carbon disulfide
Acetic acid
Methylene chloride
Trichloroethylene
Benzaldehyde
Isopropanol
Crtsol
Ethyl aery late
Monoethanol ami ne
Acetone
Methyl ethyl ketone
Naphtha
Lacquer solvent
Gasoline
Kerosene
Vinyl acetate
Sulfuric acid
Methylene chloride
Methyl ethyl ketone
Carbon disulfide
Cresol
Monethanolamine
Lacquer solvent
Sodium hydroxide (5N)
Gasoline
N , N-D1 methyl f ormami de
26
25
25
25
25
25
20
25
25
25
20
25
20
25
20
26
20
2!>
25
20
25
0
0
0
-5
-5
-5
-5
-5
-5
-5
ication Conditions
Substrate Appl
Condition
dry
dry
dry
wet
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
dry
wet
dry
dry
dry
dry
wet
dry
dry
wet

ication Rate
(g/cm2)
0.06
0.02
0.02
0.02
0.04
0.04
0.02
0.09
0.06
0.08
0.02
0.04
0.05
0.04
0.05
0.06
0.06
0.04
0.04
0.06
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
Total Liquid
Transmission
iml/24 hr)
6.9
153
377
7.0
1.2
28.7
3.5
29.6
2.0
9.4
6.6
2.8
11.7
1516
15.6
5.5
3.5
10.9
7.7
2.0
13.6
> 1
20.1
46.7
9.7
2.5
4.6
8.8
100
30.3
[~45 ml/hr]
    A secondary objective of these tests was to determine some general
trends in the comparison of softening points, viscosities, and
composition.  Such a comparison was possible since the polymers are
mutually miscible, forming mixtures with intermediate properties.

    Three tests we made using combinations of the three  polymers which had
softening points between 90 and 105°C.  The mixtures were placed in  a
steel cylinder, heated and shaken to form a homogeneous  liquid, pressurized

                                      56

-------
                TABLE A-10.  SELECTED PROPERTIES OF THREE ALLIED
                             CHEMICAL COMPANY POLYETHYLENE POLYMRS
Product
Designation
AC-1702
AC-617
AC-715
Softening
Point
850C
1020C
1090C
Density
0.88 g/ml
0.91 g/ml
0.92 g/ml
Melt Viscosity
@ 14QQC
40 cps
180 cps
4000 cps
with nitrogen, and sprayed through several types of nozzles in an effort to
obtain a continuous film.  The spraying tests were conducted at polymer
temperatures of 140°C, 1250C, and 1500C.  None of these mixtures
formed a thin film (<3 mm) and all films were susceptible to gaps forming
between successive spray passes.  In addition, thermal contraction raised
the edges of the coatings and caused severe cracking  in those which formed
the better films.  None of the final samples was capable of retaining a
liquid head, although application rates were between  0.37 amd 0.8 g/cm^.

Low-Temperature Curing Studies

    Those sealing materials which showed acceptable performance in terms of
chemical compatibility and permeability were farther  examined to assess
their ability to form a continuous film under adverse conditions. Accidents
usually occur during inclement weather when deployment of the sealing
material would be difficult; cold and wet soil substrates were considered
to be the worst conditions for establishing a sealed  surface against
hazardous materials.

    Test films of Borden Chemical's  Polyco 2607  latex, Ashland Chemical's
EP 65-68/88 urethane, and Callery Chemical's 115 urethar.e were prepared on
"dry" and "wet" substratos and at selected test  temperatures.  The Teflon
test apparatus and sand substrate were maintained  for at  least 1 hr at the
test temperature and then placed on  the conveyor,  sprayed,  and returned to
the test environment to observe the  curing process.   The  substrate
temperature and application  rate were  the controlled  variables.

Results with Polyco 2607 —
    Data obtained with Polyco  2607  are  summarized  in  Table  A-11.   The
"apparent cure time" was measured as the  time required to obtain  a
non-tacky surface film.  Using a  15  to  30 min cure time  as  the cut-off
point for the proposed use,  a  dry substrate  temperature  of  about  7  to
10°C  (45 to 500F)  is  the  lower limit for  this candidate  sealing  system.

    Some of the coatings prepared  in this series of tests were used  1n
chemical compatibility  tests.  These tests,  comparing the chemical
                                      57

-------
compatibility of coatings prepared at these lower temperatures with coatings
prepared at room temperature, are included in Table A-5.  The results of
these exposure tests were similar to those at room temperature.  The total
liquid transmitted through the coating formed at low temperatures was simi-
lar to that measured in previous tests, indicating that the coating, once
formed, is similar to that prepared at room temperature.

Results with Ashland's EP 65-86/88 —
    The lew temperature curing tests for Ashland's EP 65-86/88 ere
summarized, in Table A-12.

    A surface film formed on the Ashland urethane when the substrate
temperature was as low as -180C.  The curing process was considerably
slower at these temperatures, but part of the apparent lag could be
attributed to the amount of catalyst used.  Previous Ashland sarcp1^ cured
after approximately 1 min. at room temperature while this sample required
nearly 2 min. under the same conditions.  Ashland assured us that small
increases in catalyst concentration would promote low temperature curing
without affecting the 6 to 9 month shelf life of this formulation.

    Compatibility tests of coatings prepared at these lower temperatures
are compared with those prepared at room temperature in Table A-9.  The
results show no general trends differentiating coatings prepared on cold
substrates from those prepared on warmer substrates.  Only slight
variations were noted in most cases, except that the cold-cured urethane
appeared to withstand sulfuric acid, nonoethanolamine, and sodium hydroxide
somewhat better.  This increased resistance may he attributed to the
influence of longer curing periods rather than the cure temperature.  The
time  interval between preparation and exposure to challenge liquids was
measured in hours for samples prepared at lower temperatures compared to
minutes for samples prepared at room temperatures.

Results with Callery's 115 —
    Gallery Resin 115 was observed to form surface films on the sand
substrate at temperatures as low as -20°C (-40F)  (Table A-13).  These
films  required considerably  longer times to form  than at normal cure
temperatures and showed no evidence of foaming.  The cure resulted  in a
pliable film with a smooth glassy surface.

    The lower temperatures significantly retard the rate of cure.   Water
present in the substrate  reacts exothermally with the isocy*nat,e to promote
further reaction» but does not  interfere with the production of  an
impervious urethane coating.  Samples which were  kept at -20  to  -25°C
(-4 to -13°F) foamed  as they were allowed to warm to room temperature.
this  would indicate that  the components had not reacted completely  and only
a  surface film was formed.

    Compatibility tests of the  lower-temperature-cured  films  are compared
with  those for foams  prepared at room  temperature in Table A-8.  These

                                      58

-------
TABLE A-ll.   LOW TEMPERATURE  CURING OF 60RDEN CHEMICAL
            COMPANY  "POLYCO  2:607"
Condition of Temperature Range
Substrate (°C) (°F)
Dry 21
17
7
4

-15
to
to
to
to
0
to
23
20
10
5

-12
70
62
45
39

5
-18
Wet 21
16
7
4
0
to
to
to
to
to
23
18
10
5
4
70
60
45
39
32
to
to
to
to
32
to
0
to
to
to
to
to
74
68
50
42

10

74
65
50
42
39
Apparent Cure Time
(min)
1 to
1 to
15 to
30 to
40 tc
15 to
>120
15 to
25 to
40 to
50 to
120 to more
15
7
50
50
50
40

20
40
50
120
than 180
 TABLE  A-12.   LOW  TEMPERATURE  CURING  OF  ASHLAND CHEMICAL
             COMPANY URETHANE FOAM "EP 65-86/88"
Condition of Temperature Range
Substrate (°C) (<*)
Dry 0 to 5
-2 to -5
-6 to -10
-10 to -15
-15 to -20
Wet -2 to -5
-5 to -10
-6 to -10
-10 to -15
-15 to -20
32 to 41
23 to 28
14 to 21
5 to 14
-4 to 5
23 to 28
14 to 23
14 to 24
5 to 14
-4 to 5
Apparent Cure Time
(min)
15 to 30
25 to 60
40 to 80
55 to 80
70 to 90
35 to 40
45 to 95
70 to 90
130 to 150
120 to 140
                            59

-------
tests were done after the bottom half of the Teflon apparatus was removed
from the refrigerated environment and permitted to equilibrate at room
temperature for approximately 1 hr.  The dome was placed over the coating
and the liquid head established in tiie burette, exactly as had been done in
?rovio-.:5 tests.
           TABLE A-13.  LUW TEMPERATURE CURINb OF CALLEKY CHEMICAL
                        COMPANY UKETHANE "115" FORMULATION
Condition of Temperature Range
Substrate (°Cj
Ory 0
-6
-15
-20
to
to
to
to
1
2
-2
-10
-16
Apparent Cure Time
(OF)
32
21
5
-4
to
to
to
to
35
28
10
4
5
7
15
12
(min)
to
to
to
to
15
20
20
20
    Wet
  0  to   2     32  to   35
 -6  to  -2     21  to   28
-15  to  -10    5  to   10
-20  to  -16   -4  to    4
20 to  25
10 to  25
15 to  20
15 to  31
    Only acetone  and  N,N-dimethylformamide resulted  in  failure of  the
coatinq during a  24-hr exposure.   In  some cases  the  liquid  transmission was
slightly less  than  observed  with  a  coating prepared  at  room temperature.
In otners,  it was slightly greater.   The chemical  compatibility of the
Gallery Chemical  Company  urethane prepared at  -low  temperatures was not
significantly different from that of  the normally  prepared  urethane.

EQUIPMENT

Application  Systems

    The selection and evaluation  of equipment  to be  used  in applying  the
candidate  sealing materials  to  the  spill site  was  limited to commercial
components,  just  as was done with the sealing  materials.

    Applicator systems which were considered  included:   1)  hand-pump
pressurized; 2} propellant pressurized;  and  3} powered-pump pressurized.
These  three  general types of systems  were evaluated  for:

    1) simplicity;
    2) maintenance;
    3) logistical support; and
    4) economy.

                                      60

-------
    Data from our experimental work indicated that between 80 and 160 Ib of
material may be required to seal a spill containment area of 1200 t't^.
It was assumed that this area should Le covered in 1'j to 3u ("in.  The
probable range for dispensing rates, using these assumptions, falls in the
range of 2,7 to 10.7 lb/inin.

    One other parameter which does not appear in the experimental data was
used in the selection of sealing materials.  A maximum  liquid viscosity of
approximately 20JO cps was considered in the selection  of candidate sealing
and application systems, since the power requirements for the applicator
systems are directly proportional to the viscosity.

Hand-Pumped Systems —
    Hand-pumped systems, such as those used for insecticide and paint
spraying, may be used to apply these sealing materials.  Units are available
with capacities between $.7 an?  11.4 liters (1.5 and 3  oal).  The largest
of these would provide a system  weight on  the order of  18 kg (40 lb).

    Although this type of system is attractive  in  terms of weight and
simplicity, pressure variations  and the necessity  for continuous or
intermittent pumping would be serious limitations.

Propellant-Pressurized Systems --
    Propellant-pressurized systems such as fire extinguishers, paint
sprays, and other aerosol devices are simple, self-contained units which
offer  low maintenance requirements and portability; however, detailed
design and judicious selection of components are required to achieve  ell  of
these  attributes  in a single  applicator system.

    Throw-away relf-contained, portable propellent-pressurized systems  are
commercially available for urcthanes.   Some will dispense as much as  11.5
kg (2b  lb) of urethane at rates  between 1.25 and  1.4 kg/min  (2.7f> to  3.1
Ib/min).  Although  they offer a  very simple and compact system,  the
throw-away package  is not the favored economic  approach.

    Most propellant-pressurized  systems would be capable of  delivering  l.?5
to 4.9 kg/min  (2.7  to 10.7  Ib/min).  however, the  gas capacities would  be
insufficient to provide atomizing air,  if  it was reguired,  and would  be
limited to  that required to propel  the  coating  material from the tank.

    An estimate of  the cylinder  sues required  for supplying atomizing  air
have been made on the assumption that 1t would  require  approximately
28.3 L/n>1n  (1 ft3/min) of qas for each  0.5 kg/min  (1  Ib/min) of  liquid
flow  (Table A-14).  Propellant-pressurized systems far  exceed  the  18  kg (40
lb) weight  target  1f gas cylinders  are  employed as the  propellant source.
They are favored, however,  over  hand-pumped systems  for ease of  operation
and are more easily maintained  than  power-pressurized systems.

Power-Pressurized Systems --
         Power-pressurized  systems  may  be  used  where weight  a.id  space
requirements permit for most  spraying or  other  liquid transfer
operations.  These  systems  are  of  two basic types:  electric motor  or

-------
combustion enoine-powered.  Both would require .a pufp to be included in the
system to propel the liquids and/or compress  air und, b«»c*'..se of trie extra
wtMqht and added rraintenance,  are  less attractive than  the
propel Idf-t -.;ressurized system.

    The sire arid power requirements for  these systems,  based on the arour.to
of materials to be dispersed,  their physical  properties, ar.d the Oev.red
ipplication rates, have been calculated  as  f Anplicatinn Systems

    Two possibilities were  selected as a means  of estimating equ KW (2  hp).

                                      62

-------
                TAiLE A-14.   PRELIMINARY ESTIMATES A«r CYLINDER
                               SIZES FOR SUPPLYING ATOMIZING AIR
Sealant to be           Liquid Flow
Dispensed (lt>)         Rate Ob/win)
                                             Gas Flow
                                            Rate  (cfm)
                                               Cyl1nder
                                             Description
      RO
      80
      80

     160
     160
     160
                        2.7
                        «5.3
                       10.7

                        2.7
                        5.3
                       10.7
2.7
S.3
10.7
2.7
5.3
10.7
Size 2 Cylinder
8 1n. x ?7 In.
Tare Wt. 78 Ib
Size 15 Cylinder


           tWI.f  A. 15.
                                    Pf>««l
                                    CC»UW.
                                                    *MO BATTKt
 Ou«nf.1ty
of Co*tln
-------
    The operating procedure for  the system defined  in  Case  I would be
similar to that used for the VSA Kigi-Pak  loO package.  Additional
operations would  include cori^ction to  the manifold, connection of the
CD? cylinder, and pressurization.  First  estimates  of  Uie additional
Height and cost have also or-en n.ade.   The  proposed  system would  involve  a
•«eib:it penalty of about 4.6 k>) (10 It))  compared  witii Lit? present  kigi-Pak
liij system.  Approximately  V50 additional  in material  costs would be
incurred  iri preparing  the proposed system.

    Operating procedures for the equipment evaluated  in Case II would  be
similar to those  presently  used  by fire companies.   The base system,
including power source, would be located  at  a safe  distance but  convenient
for transporting  the urethane containers  and spray  gun to  the  spill  site.
The nrpthane package would  be the same  as in Case  I, except for  a solvent
container nestled  in the cusp between  the component reservoirs.

    Weight penalties for this system would result  from the  drag  of the hose
and extra weight  of  a  rouseable  gun.   These  are  roughly  estimated to be
several pounds.   The major  weight penalty for  power source  and compressor
are estimated to  be  12b  to  136 kg  (2/S  to 300  1b).

    The non-recurring  equipment  cost  for the components  included in  this
type  of system  are estimated  to  be  $1800.  The  cost penalty to the urethane
dispensing system, excluding  the power  source.  Is  estimated to be In the
range of  5600 to  ilOGU.


PILOT-SCALE  TESTS

    Data  provided by laboratory  tests  with the Teflon  permeation apparatus
were  used to develop a te
-------
Cardboard
Sampling Ared
                                                                       Substrate
          Structure to
          Position and
          Seal  Sides
                        Figure A-4.  Modified  test apparatus.

-------
Irregularities wvre weII-documented in each test.

    The pilot-scale tests wore designed to use water as the challenge
aqt-nt.  The compatibility and permeability tests  indicated that tests with
other challenge agents were unnecessary to determine the operating and
sealing capabilities of urethane systems.  Wato*" was p'uced over the
surface film to a depth of 30.b CM  (1 ft) and monitored throughout the
24-hr tests with a "U"-tube manometer.

    The initial tests were made with  local soil  placed  in the  apparatus  to
forty a level substrate.  After determining that  this local soil surface
could be sealed, stones and qravel. sand, water,  and qrass wrre included in
the test substrate to measure their influence on  the formation of a
continuous seal.

    The majority of these tests were  made with Gallery  Chemical Company's
basic urethane foam formulation.  Several tests  also were done with Ashland
Chemical Company's urethane to verify tne similiarity of ureUiane
(emulations.

    The kigi-Pak portaole unit, designed primarily  for  mine use, was  used
for application of the foams.  This system needs  no external  sourc*1 of  air
but mixes the components and  a volatile fluorocarbori foam agent with  an
in-line helical mixer.  The basic urethane formulations xere  applied  to  the
test substrates by spraying through a Binks Model  13 spray gun, a hano-held
version of the automatic un>t used  in the spray  chamber.  Components  were
added from two reservoirs by  nitrogen pressure,  which was regulated to
supply the proper ration. Atcmiiing air was also supplied from the
laboratory  air  line to run the mixer  and establish  the  spray  pattern.

    The apparatus ust-d  in the pilot-scale tests  (Figure A-4)  was  prepared
for spraying  the candidate sealing  systems with  a piece of cardboard
covering the  sealing channel.  After  the urethane seal  was applied,  this
cardboard was cut free  and the sides  installed.   The final seal  to  the  side
boards was made with another  frothing i-rethane  formulation, which
eliminated any chances  for leaks  at the edges of the sample.   Several
sample areas  were cut frcwi the cardboard and  the application  rate was
measured by  their vslght gain.

    The results of the  pilot-scale  tests of urethane sealing  systems  are
collected  in  Table n-16.  Sealing  soil  substrates with  urethane foams
required application of  approximately 0.1  g/cm2  (0.2  Ib/ft2).  Similar
application  rates would  seal  substrates  containing small  stones and gravel,
but not reproducibly.   The precision  measured  for application rates was
approximately 0.01 g/cm2 throughout the  coating.  The  precision In
preparing  seals which would  survive the  ?4-hr  sealing  test was less tian
50X,  however.

    The nllot scale  testr. showed  that experience Is a  key factor 1n
applying coatings  under  the  variety of  conditions expected and In
understanding the  capabilities  and  limitations  of these foamed coatings.

                                      66

-------
TABLE A-16.  RESULTS OF PlLUT-SCALE TESTS USINb URETHANE SEALANTS
Gallery Chemical KS Urethane
Application
Substrate Rate (g/cm?) Performance
Sieved Soil <1 cm
«3/8 in.) dia.





Stones & Gravel




Sod {evenly cut)


Growing Grass
(uncut)

0.05
0.12
0.25
0.11
0.17
0.1
0.08
0.13
0.49
0.13
0.17

0.14
0.17

0.5
0.55

Failure
"
•
•
Pass
•
•
Failure
M
Pass
N

Failure
N

Failure
*

Callery Chemical Rigi-Pak Ashland Chemical £P 65-86/S8
Application Application
Kate (g/c.'~2) Performance Rate (q/cm^) Performance
0.05
0.1
O.U
0.13
0.12


0.09
0.16
0.13
0.14
0.13
0.16
0.15
0.14
0.16
.22
0.53
Failure
•
•
Pass
N


Failure
IS
•
Pass
M
Failure 0.7 Failure
tt
Pass
Failure 0.6 Failure
M
•

-------
The bridging ett'ect of the foaming action, which is so necessary to make an
effective seal with a minimum of sealant, depends heavily on such va;  iables
as sealant and substrate temperatures,  the presence of water, the appli-
cation rate Ob/min), and the manner  (sinqle or multiple passes) in which
it is applied.  Values could be provided  for each condition, but only
experience will provide the necessary  "feel" for people in the  field where
all these variables may be changing simultaneously.

    The results of the pilot-scale tests  also  indicated that the influences
of wind, surface area, and surface defects are difficult to measure,  and
compensating means are difficult to describe to  inpy.oer ienced personnel.
Whether tne unframed  film coverage rate will form a continuous  coating  is a
judgement mac*e by the person applying,  it.  The  operator must estimate  the
influence each of these factors would  have and modify  the application
technigue accordingly.

    Comparing  frothed and normal foaming  formulations  of commerical ure-
thane systems  showed no specific advantages for  either type-.  Urethane
foams appeared to have  less penetration into normal substrates, while  the
froths appeared to bridge larger gaps  and debris.   Neither  showed  a
definite advantage on vegetated substrates.  Urethane  froths, however,
appeared to be somewhat more effective for application on cold  or  wet
substrates.

    The course of the program  was  altered when u-etuane foam was used in
attempts to seal grass-covered  substrates.   In these  tests,  coverage  rates
in excess of  0.6 g/crn^  (1.3  lb/ft^)  proved incapable  of sealing these
substrates against water.  With  a  design  goal  of scaling  1200  ft2,  the
coverage rates experienced,  even  if  successful  in  sealing,  precluded  its
use as  an emergency  portable  system  for grassy substrates.   This limitation
was considered too severe for  the  anticipated  use  of  the  sealant system to
continue further development.

    The basic problem areas  appear to be 1n  the  "shadow effect," encoun-
tered during  spraying,  and  in  the  nature of  the  foaming process.  "Shado-
wing"  is common  to any  spray application and  reguires exceptional  care and
conscientious application from all  angles to  assume complete coverage of
the surface.   The  foaming process  for the urethane formulations used in
this  program  helped  f.o  minimize the  prnblems  of  bridging surface
discontinuities  and  "shadowing" 
-------
in conjunction with inclement weather severely restricts the application of
the urethane systems.  This limitation \
-------
                                     APPENDIX  B
                 English  to  Metric Conversion Factors
      t*«llib tmit
                                       KulttplUr
                                                                      H.UIC Mil
•crv-fiwt
Cl*lc  f*»t
cukic  lick

cubic  xi r4
fMt Mr
IlllMi Mr ««n Mr *»T
|<1IM> Mr ciplti Mr *if

|>ll MI Mr 4ir M'
Mjiuri t«*l
111 loot Mr i
>4»ir« («at
tilliat MT i«t>*ri CMC
•Illlm i«i IMI Mr
•HUM |«1U« fff tff

•H*
M«« M' kill Ira
Mrtt Mr OI1IIM
                          Ct
                          cf.
            I.40S
            l.III.f
            II. U
            I.OIt]
            lt.il
                          If
                          It
                          ill.
            «.1«S
            «.»>
            l.TIS
            *.ISI
(H/M (t    I.Ml • II'
            t.ltt
            9. tit

            41.741
            I.T4<>
            1.14
            1.14
            J.1I1
            i.m.i
            41.1*1
            • .I4M
IMf

U.
U./kr
•II i«l.
                                      ».4i4

                                   >r  • 
                                rur*  l.Ul

          l.tl> inkle IM(  Ik/t.iM c(  14 «ir
         ' OilltH |«I1«W  Ik/Bll |«1.  S.Ut
                i«*t
       Mr
Mmr* mil*

       i Cukic
                            I !••
                            I •!
                                      11.11
                                      4.Ml * It'*
            ».4ii
            I. Ml
            *.IM
                         1
                         c« «/•>•
                         1/rK
                         cu c*
                         1
                         Cll •
                         I
                         *•( t-

                         •/••c

                         1
k*ctlr«
cwklc aitvr
Itt.t
citklc Mt«r* Mr vlMif*
linri Mr i«cM4
C«klC ClfttlBltlf
Ut.t
cubic «*t«r
                                                                ltt*r> Mr 4*7 Mr toctar*
                                     llt«Ti Mr »c
                                     cukic Mt«n *
                         cw/ii/kr/M • cukic s*<0r» Mr iM^r
                         l/i*c/i^ •   itl«ri Mr ••coii4l Mr

                         1/M a       lliifft Mr «4tw«r« actcr
                                                   HI
                                                   en •
                                                                c«iic j*#l
                                      Million  (Uteri i  l« )
                                      CwikC IMtlfl
                         en */lir/K •  cokic miri Mr kwr

                         l/i«c        Ittiri Mr  I*C«M|



                         •I/I         •llHi.-««i  Mr ltt*r
                                      kll«|rui
                                      kvcuri
                         I/CD •
                         •I/I
                                                   llf ftf
                                             '  caklc a»ti>r
                                             ii M* llt«r
                                             i  per cvklc m*t
                                                                (ruti»«l»r
                                                                ktt**ri«i »
rut
                                      1,117
                                      1.114
                                                    ••trie IM   otrlc >•»
                                             70
                                                             Reproduced  from
                                                             beil  available copy.
                                                                  T-'i'^.O GUIDE i--V!'I!;r

-------