EPA
         United States
         Environmental Protection
         Agency
            Office of Research and
            Development
            Washington, DC 20460
EP/V540/5-90/005a
August 1990
         Superfund
Emerging Technologies:

Bio-Recovery Systems
Removal and Recovery of
Metal Ions from
Groundwater

-------

-------
                                     EPA/540/5-90/005a
                                        August 1990
          Emerging Technologies:

    Bio-Recovery Systems Removal and
Recovery of Metal  Ions from Groundwater
             Risk Reduction Engineering Laboratory
             Office of Research and Development
             U.S. Environmental Protection Agency
                 Cincinnati, Ohio 45268

-------
                              DISCLAIMER
     The information in this document has been funded in part by the  United States
Environmental Protection Agency under Cooperative Agreement  No. CR-815318010
to  Bio-Recovery Systems, Inc.  The document has been subjected to  the Agency's
administrative and peer review  and has  been approved for publication as an EPA
document.  Mention of trade names or commercial products does  not  constitute
endorsement or recommendation for use.

-------
                                   FOREWORD
       The  U.S.  Environmental Protection Agency (EPA) is  charged by  Congress with
protecting the  Nation's  land,  air,  and water  resources.   As  the  enforcer  of  national
environmental laws, the EPA strives to balance human activities and the ability of natural
systems to support and nurture life.  A key part of the EPA's effort is its research into our
environmental problems to find new and innovative solutions.

       The  Risk  Reduction Engineering Laboratory (RREL) is  responsible for  planning,
implementing, and managing research, development, and demonstration programs to provide
an  authoritative,  defensible engineering  basis  in  support of the policies,  programs,  and
regulations  of  the  EPA with  respect to  drinking water, wastewater, pesticides,  toxic
substances, solid and hazardous wastes, and Superfund-related activities.  This publication
is one of the products of that research and  provides a vital communication link between the
researcher and the user community.

       Now  in its fourth year, the Superfund Innovative Technology Evaluation (SITE)
Program is part of EPA's research  into cleanup methods for hazardous waste sites around the
nation.  Through cooperative agreements with developers,  alternative  or innovative
technologies are  refined at the  bench-and pilot-scale level and then demonstrated  at actual
sites.  EPA  collects and evaluates extensive performance  data on each technology  to use in
remediation  decision-making for hazardous waste sites.

       This report documents the results of laboratory and pilot-scale field testing of dead,
immobilized  algal cells in a silica gel polymer to remove heavy metal ions from  mercury-
contaminated groundwaters.  It is  the first in a series of reports sponsored by the SITE
Emerging Technologies Program.


                                         E. Timothy  Oppelt, Director
                                         Risk Reduction Engineering Laboratory
                                         iii

-------
                                  ABSTRACT
       A series of laboratory tests and an on-site pilot scale demonstration of Bio-Recovery
Systems'  AlgaSORB® technology for the removal and recovery of mercury-contaminated
groundwaters were conducted under the SITE program.

       Optimum conditions were determined for mercury binding to AlgaSORB®. Conditions
under which mercury could be stripped from AlgaSORB® were also developed.

       On-site, pilot  scale  demonstrations  with  a portable waste  treatment  system
incorporating  columns  containing two  different  AlgaSORB® preparations confirmed
laboratory tests.   Over 500 bed volumes of mercury-contaminated groundwater could be
successfully treated before regeneration of the system was required.  Mercury was removed
to levels below the discharge limit of 10 u.g/L.

       This report was  submitted in  fulfillment  of  Cooperative Agreement Number CR
815318010 by Bio-Recovery  Systems,  Inc.  under  the  partial sponsorship of the U.S.
Environmental Protection Agency.  This report  covers a  period from October,  1988 to
January 31,1990,  and work was completed as of January 31,  1990.
                                      iv

-------
                           TABLE  OF CONTENTS
                                                                            Page
     Disclaimer	:'.
     Foreword	'.'
     Abstract	IV
     Figures	v
     Tables	•	i	-v.
     Acknowledgements	v"

   I. Executive  Summary	^
  II. Introduction.	|
 III. Conclusions and Recommendations	•	«*
 IV. Background  Information	•	j1
     A.  AlgaSORB® Description and Previous Work	4
         1.  Introduction	4
         2.  Waste Streams for Which the AlgaSORB® and Other Ion Exchange   •
            Technology Is Applicable	6
     B.  The Use of AlgaSORB® and Ion Exchange to Effect Heavy Metal Waste
         Minimization:  Comparison to Conventional  Waste Treatment	6
     C.  State of Development	8
     D.  Application of AlgaSORB® to Metal-Contaminated Groundwaters and
         Wastewaters	] 'P
         1.  Removal of Cadmium from  Water at a Superfund Site	1 0
         2.  Removal of Copper from Contaminated Groundwaters Containing
            Halogenated Hydrocarbons	1 °
         3.  Removal of Mercury from Contaminated Groundwaters..	1 1
         4.  Selective Removal of Lead  from Wastewaters	-	1 1
   V. Description of Site Containing Mercury Contaminated Groundwaters	1 3
  VI. Laboratory Testing	,	•	•	•	1 j>
     A.  Experimental  Procedures	1 5
     B.  Results	•	\ °
         1. Water Analysis	•	1 6
         2  AlgaSORB® Tests	1 7
 VII. On-Site,  Pilot  Scale Demonstration	30
VIII. Quality Assurance	•	3^
     A.  Verification of Modification of  EPA Method 245.1 for Mercury Analysis	35
     B.  Analysis of EPA-Provided Standard	36
     C.  Mercury Spikes	•	3 ^
      D.  Mercury Analysis in the Presence of Thiosulfate	40
      E.  Analysis of Samples Resulting from On-Site Testing	•	41

-------
                              LIST  OF FIGURES
Number
   1.  Recycle-Recovery System	8
   2.  Automatic Recycle-Recovery	9
   3.  Portable Wastewater Treatment System Used for On-Site Testing...:	3 1
                               LIST OF TABLES
Number
  1.  Average Composition of Mercury-Containing Groundwaters	..13
  2.  Seasonal Variation of Mercury Concentration in Monitoring Wells	1 4
  3.  Mercury Concentration in Groundwaters	1 6
  4.  Analysis of Effluents from a Column Packed with AlgaSORB®-602	.1 7
  5.  Analysis of Stripping Effluents from Column Loaded in Table 4	 1 8
  6..  Analysis of Effluents from a Column Packed with AlgaSORB®-602	!l 8
  7.  Analysis of Effluents from a Column Packed with AlgaSORB®-602	1 9
  8.  Analysis of Effluents from a Column Packed with AlgaSORB®-602	20
  9.  Analysis of Stripping Effluents from Column Loaded in Table 8	!!."!.'.'20
 10.  Analysis of Effluents from a Column packed with AlgaSORB®-602	2 1
 11.  Analysis of Stripping Effluents from Column Loaded in Table 10	..".'.....22
 12.  Analysis of Effluents from a Column Packed with AlgaSORB®-601	2 3
 13.  Analysis of Stripping Effluents from Column Loaded in Table 12........	24
 14.  Analysis of Effluents from a Column Packed with AlgaSORB®-603	..25
 15.  Analyses of Stripping Effluents from Column Loaded in Table 14	2 6
 16.  Analysis of Effluents from a Column Packed with AlgaSORB®-602	.......2 6
 17.  Analysis of Effluents from a Column Packed with AlgaSORB®-603	2 7
 18.  Analysis of Effluents from Two Columns in Series Packed with
      AlgaSORB®-624 and AlgaSORB®-640	2 8
 19.  Analysis of Effluents from Two Columns in Series Packed with
      AlgaSORB®-624 and AlgaSORB®-640	2 9
 20.  Variation in Mercury Content of Groundwaters During On-Site Pilot
      Scale Testing	           32
 21.   On-Site Pilot Testing for Mercury  Removal from Groundwaters	.....33
 22.  Analysis of Effluents from AlgaSORB®-624 Column on the Portable
     Treatment System	34
 23.  Mercury Analysis of Standards Using Sodium Borohydride as a Reductant	.".35
24.  Mercury Analysis of Standards Using Sodium Borohydride as a Reductant	:....3 6
25.  EPA-Provided Sample Information	     37
                                      vi

-------
26.  Mercury Analysis of EPA Water Pollution Quality Control Sample	38
27.  Error and Recovery Analysis of Mercury Spikes	•	39
28.  Effect of Thiosulfate on Mercury Analysis	•	40
29.  Analysis of Mercury-Thiosulfate Samples Oxidized  with Hydrogen Peroxide	40
30.  Mercury Analyses of Thiosulfate-Containing Solutions Without Acid Digestion	41
31.  Identification of Samples Sent to Woodward-Clyde Consultants and EER
     Technologies for Mercury Analysis..	..;.....,..	4 2
                                        vii

-------
                            ACKNOWLEDGEMENTS
       This document was prepared under Cooperative Agreement No. CR 815318010 by
Bio-Recovery  Systems,  Inc.,  Las Cruces,  NM  under the  sponsorship of the U. S.
Environmental Protection Agency.  Naomi  P. Barkley of the Risk Reduction Engineering
Laboratory, Cincinnati, Ohio was the Project Officer responsible for the preparation of this
document and deserves  special  thanks  for her helpful comments and advice.  Special
acknowledgement is given to Donald E. Sanning, Chief, Emerging Technology Section, SITE
Demonstration and  Evaluation  Branch, Superfund Technology Demonstration Division for
providing technical guidance and input.

       Participating  in the development of this report for Bio-Recovery Systems, Inc. were
Dr. Dennis W. Darnall and Michael Hosea.  Special recognition is given to Sandy Svec, Dr.
Maria Alvarez, Rafael Tamez and  David Marrs for laboratory and on-site pilot testing and
coordination of analysis.
                                     viii

-------
                              I. EXECUTIVE SUMMARY
       During  1989 laboratory and on-site  pilot scale testing of Bio-Recovery Systems'
AlgaSORB® technology for  the  removal and  recovery  of  mercury from contaminated
groundwaters were conducted. AlgaSORB®, a non-living,  immobilized algal bio-mass, was
packed into columns through which the mercury-contaminated groundwaters were pumped.
Mercury  concentrations  in  influent and  effluent  were  measured to  determine the
effectiveness of mercury removal.  Once the columns showed unacceptable mercury leakage
(10 u,g/L), the columns were stripped of mercury and reused.

       Several  different AlgaSORB®  preparations  containing different algal species were
tested for effectiveness in mercury removal.

Summary Results

       AlgaSORB® testing was complicated by the fact that over the sampling period mercury
concentrations in the groundwaters varied by over an order of magnitude from 150 u.g/L to
1550
       In addition it was found that one variety of AlgaSORB®  showed  varied mercury-
binding capability with  waters  collected at various times.  This  suggested a variation  in
mercury speciation over the sampling period.  Because of these variations, final on-site
pilot scale testing was done with a blend of two AlgaSORB® preparations.  One preparation
had a rather high mercury capacity but also exhibited a rather  high leakage of mercury and
the second preparation  had a lower mercury binding  capacity  but exhibited low leakage  of
mercury.

       On-site,  pilot scale testing was conducted November  7 to December 1, 1989.  A
portable water treatment system that contained columns of the two different AlgaSORB®
preparations was tested over the three week period.  Waters were pumped through the
AlgaSORB®  resins at  a flow rate of  6 bed volumes per hour.  Over 500 bed volumes  of
mercury contaminated  waters  were passed  through the resins before effluent mercury
concentration  exceeded discharge levels of  1.0 |xg/L.  These results suggest that  a full-scale
treatment  system would be effective for  mercury  removal  from  groundwaters.   Costs
associated with such  a  treatment system should be  typical of those  associated  with
commercial ion  exchange systems for treatment of industrial waste waters.  In  contrast  to
commercial ion exchange resins, however, AlgaSORB® functions well with waters which
have a high total dissolved solid  content and which contain organic compounds.

-------
                                 II. INTRODUCTION
       The Superfund Amendments and Reauthorization Act of 1986 (SARA) directed the
 Environmental Protection  Agency (EPA)  to establish  an  "Alternative  or Innovative
 Treatment Technology Research and Demonstration Program."  In response, the EPA's Office
 of Solid  Waste and  Emergency  Response and the Office of Research  and Development
 established a formal  program called the Superfund Innovative Technology Evaluation (SITE)
 Program, to accelerate  the development and use of innovative cleanup technologies at
 hazardous waste sites across the country.

       The SITE Program is comprised of the following five component programs:

              Demonstration Program
              Emerging Technologies  Program
              Measurement and Monitoring Technologies Development Program
              Innovative Technologies Program
              Technology Transfer Program

       This  report is the  first in a series of reports  sponsored by the SITE  Emerging
 Technologies Program.  Before a technology can be accepted into the Emerging Technology
 Program, sufficient data  must be available to validate its  basic concepts.  The technology is
 then subjected to a combination of bench- and pilot-scale testing in an attempt to apply the
 concept under  controlled conditions.

       Bench- and  pilot-scale  testing of the Bio-Recovery Systems,  Inc.  AlgaSORB®
 technology has  been performed under  the  SITE Emerging  Technology Program.  The
 AlgaSORB® technology is designed to remove heavy metals  from aqueous solution.  The
 process is based upon the natural affinity of algae cell walls for heavy metal ions.  The
 sorption  medium,  AlgaSORB®, is composed  of a  non-living  algal bio-mass which is
 immobilized  in a silica polymer.  AlgaSORB® is a hard  material which can be packed into
 columns  which, when pressurized, exhibit good flow  characteristics.  This technology is
 useful  for removing  heavy metal  ions  from groundwaters  that contain  high  levels of
dissolved solids.

       Groundwater contamination is found at over 70 percent of the sites currently on  the
 National  Priority List (1).  Groundwaters have been contaminated with either, or both,
toxic organic molecules  and heavy metal ions.  The most common means of addressing
contaminated groundwater is extraction and  treatment.   While biological  in situ  treatment
of groundwaters contaminated with organics may be possible,  there is no effective  method
for in situ treatment of  groundwaters  contaminated with heavy metals.  AlgaSORB®  was
developed for removal of dilute concentrations  of heavy metals from groundwaters. ;

-------
                      III. CONCLUSIONS AND RECOMMENDATIONS
A,     Conclusions:

       On-site, pilot scale testing of AlgaSORB® showed effective  mercury recovery from
contaminated groundwaters.  However, initial  laboratory experiments  showed the,' dangers in
making conclusions from a single groundwater sample.  These studies showed that not only
did mercury concentration vary over the  sampling period, but also the data suggested that
the chemical species of mercury varied over the sampling period as well.  In the end it was
found possible to combine two different AlgaSORB® preparations to  effect mercury removal
from groundwaters to levels below 10 u.g/L.
B.
Recommendations:
       Work  done at  the site described herein  indicates that a full treatment system for
mercury recovery can be installed.  However, because the chemistry of other groundwater
sites will undoubtedly differ from the one tested  here, laboratory treatability testing  will be
required before the technology can be applied at other mercury-contaminated groundwater
sites.

-------
                          IV. BACKGROUND INFORMATION
A.   AlgaSORB® Description and Previous Work

1.   Introduction      ,

     The use of microorganisms in the treatment of hazardous  wastes containing both
  inorganic and organic pollutants is becoming more and more common.  There have been two
  approaches to the use of microorganisms in .waste treatment.  One  involves the use of living
  organisms  and the  other  involves  the  use  of  non-viable  biomass  derived  from
  microorganisms.  While  the use  of living organisms is often successful in the treatment of
  toxic organic contaminants, living organisms have not been found  to be  useful  in the
  treatment of solutions containing heavy metal ions.  This is  because once  the  metal ion
  concentration becomes too high or  sufficient metal ions are adsorbed by the microorganism,
  metabolism is disrupted causing the organism to die. This disadvantage is not encountered if
  non-living organisms or biological materials derived from  microorganisms are  used to
  adsorb metal ions from solution.  Instead the  biomass is  treated as another reagent,  a
  surrogate ion exchange resin.  The binding, or biosorption, of metal ions by the biomass
  results from coordination of the  metal ions to various functional groups in or on the cell.
  These  chelating groups, contributed  by the cell biopolymers, include carboxyl,  imidazole,
  sulfhydryl,  amino,  phosphate, sulfate, thioether, phenol, carbonyl,  amide  and  hydroxyl
  moieties (2).

       Various algal species and cell preparations have quite different affinities for different
  metal ions  (3-4).  The different and unusual metal binding properties exhibited by different
  algae species are explained by the fact that various genera of algae have different cell wall
  compositions.  Thus, certain algal species  may  be much more effective and selective than
  others  for removing particular metal ions from aqueous solution (5).

       The reaction of heavy metal  ions with  a non-living algal qell  forms complexes which
  are composed of  the algal cell and  the metal  ions.  The result of this reaction, i.e., the
  formation of the alga-metal ion  complex  is basically why  metal ions are  toxic  to living
  organisms and explains how the toxic effect of metal ions is amplified in the food chain. The
  metal ions  are adsorbed to the cell even  at concentrations in the mg/L-u.g/L range. The
  bound  metal ions,  when accumulated over time, eventually interfere with metabolism  by
  disruption of enzyme reactions  and kill the organism.  If microorganisms on which metal
  ions have been sorbed  are used as a food source by larger organisms, the metal ions find
  their way into the food chain which  can eventually result in toxic effects for humans.

       While the interaction of metal ions with microorganisms has been  known  for many
  years, it is only recently  that advantage has been taken of the high affinity of microorganism
  cell walls to remove and  recover  metal ions from industrial  wastewater or  contaminated
  groundwaters.  Methods to reverse the reaction of metaj ion sorption have been developed so

-------
that when metal ions are  recovered from dilute solutions they can be stripped off the cell
walls in a highly concentrated form. The cells can then be reused to capture more metal ions
from dilute solutions. Conditions can also be adjusted so that only one or two types of metal
ions are adsorbed from a  solution containing  several metal ions, or a variety of metal ions
can be sorbed from solution and then they can be selectively stripped from the algal cell one
metal at a  time  (2,6).

     Bio-Recovery Systems,  Inc. has developed a  proprietary,  algal based  material,
AlgaSORB®, which can be used on a commercial basis to remove and recover heavy metal ions
from point-source industrial wastewater, contaminated groundwaters  or mining  process
streams.  AlgaSORB® functions very much  like a commercial ion  exchange resin.  It can be
packed into columns through which waters containing heavy metal ions are flushed.  The
heavy metal ions are adsorbed to AlgaSORB®  and metal-free water exits the column  for reuse
or discharge.  Once the AlgaSORB® is saturated with metal ions, the metals can be stripped
from the AlgaSORB® which is then ready for reuse.  In comparison to ion exchange resins,
however, AlgaSORB® has some distinct advantages which make it superior to ion exchange
resins  for certain applications (see below).  In other instances  ion exchange resins perform
better than AlgaSORB®. AlgaSORB® has a remarkable affinity for heavy metal ions; in some
cases the metal-binding capacity is as much as 10 percent of the dry weight of the cells. The
algae matrix is capable of concentrating heavy metal ions by a factor of many thousand-fold.

     When unadulterated algal cells are packed into columns, the cells tend to aggregate and
to form cohesive clumps through  which  it  is difficult to force  water  even under high
pressures.   However, when the cells are immobilized into  a polymeric  matrix, this
difficulty  is  alleviated.

     The  algae are killed in the immobilization process indicating that sorption  does not
require a living organism, and hence the algal matrix can be exposed, with little or no  ill
effects, to  solution conditions  which would  normally  kill  living cells.   The pores of the
polymer are  large enough to  allow free diffusion  of  ions  to the algal cells, since similar
quantities of metal  ions  are bound by  free and  immobilized cells.   The immobilization
process serves  two purposes:   (I) It  protects the  alga cells from decomposition  by other
microorganisms, (AlgaSORB® immersed in aqueous solution for over two years has  shown no
decrease  in  metal binding efficiency)  and (2)  it produces a  hard material which can be
packed  into chromatbgraphic  columns,  pressurized  and  exhibits  excellent- flow
characteristics.

     In addition to the  immobilized  algal  matrix's  usefulness for  the removal of  the
"traditional"  heavy metals  from  solution,  it  also is useful for near quantitative removal and
recovery of very low concentrations (in the parts per billion range) of precious  metals s'uch
as gold, silver, platinum and palladium (7).

     AlgaSORB® functions as a "biological" ion exchange  resin and like other ion-exchange
resins, can be recycled. Metal ions have been sorbed and stripped over  many cycles with no
noticeable loss in efficiency.  In contrast to current ion  exchange technology, however, a real
advantage  of the algal matrix is that  the  components  of  hard water  (Ca+2 and  Mg+2) or
monovalent  cations  (Na+ and  K+) do not significantly  interfere with the binding  of toxic,
heavy  metal  ions.  In fact  calcium or magnesium ion concentrations as high  as  10,000 mg/L
have little or no  effect on AlgaSORB® sorption of copper  at concentrations as low as 6.5
mg/L.  The binding of Ca+2 and Mg+2  to  ion-exchange resins (even chelating ion exchange
resins  which are  relatively selective  for transition metal ions)  often limits ion exchange

-------
usefulness since calcium and magnesium ions are frequently present in high concentrations
and compete with  heavy metal ion binding.  This means that frequent regeneration of ion-
exchange resins is necessary in order to effectively remove heavy metal ions from solutions.

     AlgaSORB®  is also effective for heavy metal removal from waters containing organic
residues.  Organics often foul synthetic ion exchange resins  which  limits  their utility  in
many wastewater treatment applications, including groundwater treatments.  AlgaSORB®, on
the other hand, functions well in waters containing organic molecules.

2.   Waste Streams for which the AlgaSORB® and Other Ion Exchange Technology is
     Applicable

     A  major source  of heavy  metal  wastes  from  industrial sources  comes from the
electroplating, metal  finishing  and  printed  circuit board  manufacturing  industries.
Wastewaters  from  these  industries  primarily  come from  rinsing operations.   The
rinsewaters will typically contain rather  low concentrations  (on  the order of  100 parts per
million)  of heavy metal ions.  Certain of these waste streams are particularly amenable  to
treatment with AlgaSORB® or ion exchange resins.  The metals can be recovered and then
either recycled back into the process or recovered for use by other industries.  In addition
AlgaSORB® may be useful for polishing  waste streams previously treated by other methods,
but which still have metal ions present at concentrations above compliance levels.

     Contaminated groundwaters and surface leachates often contain heavy metals in the low
parts  per million or even part per billion range.   The AlgaSORB® technology is  well suited
for removing  and recovering heavy metal ions from these waters, which will often contain
high  concentrations of dissolved materials which are non-toxic. Often these types of waters
will contain high concentrations of  sodium, potassium, calcium, magnesium, chloride  or
sulfate which are innocuous and for which no treatment is needed. AlgaSORB® is capable of
preferentially removing  heavy metals which are found in these streams. Toxic heavy  metal
ions  which can be  recovered with  the algal biomass include copper, nickel, uranium, lead,
mercury, cadmium, zinc,  arsenic and silver among others.

     AlgaSORB®  has  a higher affinity for precious  metal ions than any other heavy  metal
ions tested (5-6).   Thus another area in  which the AlgaSORB® technology is useful is  in the
recovery of  gold,  silver or platinum  group  metals from  mining process streams,
wastewaters resulting from mining  operations, and industrial point source wastewater.

B.   The Use of AlgaSORB® and Ion Exchange to Effect Heavy Metal Waste Minimization:
     Comparison to Conventional Waste Treatment

     The conventional  method for treating  wastewaters in  electroplating or  printed circuit
board manufacturing plants has been to commingle  all  metal-containing wastewaters which
are then sent to a central location for treatment.  Treatment methods vary depending upon
what metals are present in the stream,  but the most common treatment is precipitation  of
the metals as hydroxides.   If metal cyanide complexes are present, cyanide is usually
oxidized prior to  metal precipitation.  Likewise,  if  hexavalent  chromium is present, it  is
usually   reduced  to trivalent chromium prior  to  precipitation.   The  metal  hydroxide
precipitates are then dewatered and most commonly sent to a hazardous waste landfill.  Since
August 8, 1988, these metal-containing sludges can no longer be sent to a hazardous waste
landfill unless  they are stabilized so that the toxic metal ions cannot  be leached  from the
sludge.   A variety of agents such as Portland cement, fly ash or other pozzolanic materials

-------
can be used to stabilize the sludge, but whatever the stabilization method, the disposal costs
have increased dramatically since August 1988.  In addition both state and federal regulatory
agencies are moving  toward  the future complete ban of land disposal of metal hydroxide
sludges in any form.

     In addition to high  sludge disposal  cost, another disadvantage of the conventional
treatment  system  is the  difficulty  in  many  instances  of  reaching  effluent  metal
concentrations  low enough to  meet discharge standards.   This  is  because hard-to-treat
waters are often commingled with easy-to-treat waters thereby making all  the  wastewater
hard-to-treat.   For example,  in printed circuit  board manufacturing operations there are
typically three different  types  of copper-bearing  wastewaters  which  must be  treated:
copper  sulfate from  acid copper baths,  ammoniacal copper  from  alkaline etchers and
chelated (usually EDTA, quadrol or tartrate) copper from electroless copper baths.  Copper
sulfate responds very well to hydroxide precipitation,  but the ammonia complex of copper
and  the EDTA chelate of copper are  very difficult to treat with conventional hydroxide
precipitation.   Thus expensive chemicals such as sodium borohydride or dithiocarbamates
are added  to the entire wastewater stream in order to treat the ammoniacal and chelated
copper which usually make up only a small proportion of the total waste streams.

     When the conventional  hydroxide precipitation of metals  is  used,  usually sodium
hydroxide or lime  along  with other  reducing agents  or  flocculating agents are added to
produce the metal  hydroxide sludge.  Once the sludge is removed from the wastewater the
water is generally discharged to a sewer.  There is no opportunity for reuse or even  partial
reuse of the water because the effluent water has too many dissolved salts to be effective as a
rinsewater.  The cost of  deionizing  this water is generally much higher than  the cost of
deionizing fresh tap water and hence water reuse is generally not a viable economic option.

     Generators of toxic metal  sludges are held liable, without proof of fault, for cleaning
costs and natural resource damage at hazardous waste disposal sites at which the generator's
waste is disposed.  Therefore  if the owners of a hazardous waste dump happen to mismanage
the site so that toxics are allowed into  the environment, it is the generator who is ultimately
responsible for clean-up.  Thus  any process by which sludge can be minimized or eliminated
will reduce liability for the generator.

      Bio-Recovery Systems'  technology has  been incorporated into  an  effective recovery-
recycle approach  to wastewater treatment for the  electroplating, metal finishing  and
electronics industries. The concept is illustrated in Figure 1 for a  treatment system that
allows for  recovery of metals and recycling of process waters.  In this scheme rinsewaters
derived from  each individual plating bath are  segregated and  passed through columns
containing AlgaSORB® or specialty ion exchange resins.  Metal ions are removed  from the
rinsewaters which  can then be discharged directly or  returned to  the rinse  tanks for partial
water reuse.   Because  salts tend to build-up in  the rinsewaters, deionization of the
treatment effluent may be needed if it is to be reused in critical rinses. Otherwise  a bleed-
off of water to the sewer is adequate  to keep salt-build up at acceptable levels.  Such an
approach can often decrease water usage by 50 to 90 percent.

      Once the columns of ion exchange resins or AlgaSORB® are saturated with metals, the
metal ions can be stripped from the columns.  The concentration of the stripped metals is
approximately 10 g/L.  In  certain instances these stripped metal ions can  be added back to
the plating bath.  In  instances  where  this is not acceptable, the metal can be recovered
through electrowinning or  metalwinning.   Alternatively  the metal  ions  can be further

-------
concentrated by evaporation  and sent to one of a  number of companies which are now
established to recycle such materials. Whichever  approach  is  taken,  however,  the
elimination of sludge production results in lower  operational costs due to decrease in
chemical  costs, decrease in water  usage, elimination of  sludge disposal  costs and
minimization of future  liability.
                             RECYCLE-RECOVERY SYSTEM
                             WORK
PLATING
TANK
                                         /-^  JL
                                         jr  _x_^»_x-tx>
                                      RINSE
                                      TANK
RINSE
TANK
            •~l
                CONCENTRATE
                                                      PURIFIED
                                                      WATER
Figure 1.     Recycle-Recovery System.  Segregated rinsewaters from a plating process
are  directed through  a recovery  system where metal  ions  are recovered, and  the
rinsewaters are directed back to the rinse tanks.  The concentrated recovered metals are
sent back to the plating process tank where possible.

C.   State of Development

     Bio-Recovery  Systems  is  currently  manufacturing  arid  installing  wastewater
treatment systems for  use  in recovering  heavy metals from industrial point sources in the
electroplating and printed circuit board manufacturing industries.  Figure 2 shows one  such
system which has been designed for a printed circuit board manufacturer.  The heart of the
system  is comprised  of  columns  (B)  which  contain  the metal-adsorbing  materials.
Rinsewaters which contain only a single type of plating or etching chemistry are segregated
and plumbed to individual columns.  When the columns become saturated with metal ions, a
specific metal ion sensor signals the controller (A) to begin a regeneration cycle to  strip the
metals from the  materials in the column and to send the  stripped metal ions to  one of the
holding tanks (D).   Once regeneration is complete, the controller automatically returns the
regenerated column back  into service.   The  stripped  metals are then  recovered as the
metallic  elements in the metaiwinning unit (E).

     The system shown in Figure 2 is capable of treating 30 L/min (8  gal/min),  however
larger flow rates  (up to hundreds of gallons per minute) are accommodated by simply adding
either more  metal-adsorbing columns or by using larger diameter columns.

       The  system shown  in Figure   2 was designed for  a  printed  circuit board
manufacturer, but the same1 type of system is also employed for metal finishing  and
electroplating facilities.

-------
                                               c
B
Figure  2.    An Automatic Recycle-Recovery Wastewater Treatment System. A.  controller.
B. metal adsorbing modules. C. deionized water system/ D. holding tanks for pH adjustment,
regenerant chemicals.  E. metalwinning module.

-------
Different chemistries are encountered in  metal finishing rinsewaters, but the approach to
treatment of these waters is basically the same as that encountered in a printed circuit board
manufacturer's facility, i.e.,  wastewaters  are  segregated for treatment  so that maximum
reuse, of metals and water can occur.

D.   Application of AlgaSORB® to Metal-Contaminated Groundwaters and Wastewaters

     In 1986  and 1987  Bio-Recovery  Systems was awarded Small Business Innovative
Research (SBIR) contracts from the United States Environmental Protection Agency (EPA)
to research and develop the AlgaSORB® technology for commercial applications. Results from
these contracts, some of which are summarized below, show the efficiency of AlgaSORB® for
heavy  metal removal from a variety of  sources.   These successful laboratory tests led  to
Bio-Recovery's participation in  the SITE  program, through submission of a pre-proposal to
the Emerging Technology Program.

1.   Removal of Cadmium from Waters at  a Superfund Site

     Officials  from EPA Region II arranged to supply samples from a well at a Superfund
site in  New Jersey, the Waldick Aerospace Devices site. These waters were contaminated
with cadmium  at a level of 0.13 mg/L. The waters at a pH of 6.0-7.1  also contained, among
other organics, 0.66 mg/L of a halogenated hydrocarbon, tetrachloroethylene.

     A column containing  AlgaSORB® (0.7 cm i.d. x 13 cm high) was prepared, and the
Waldick Aerospace waters were passed through the column.  Five mL fractions of water
exiting the column were collected until 500 mL (100 bed volumes) of Waldick waters were
passed through the column at a flow rate of one-sixth of a bed volume per minute (total bed
volume was 5.0 mL).  Each fraction  of effluent  was analyzed for cadmium using  graphite
furnace atomic absorption  spectrometry.  All  effluent fractions showed that cadmium
concentration was near or below 0.001 mg/L after the passage of the 100 bed volumes of the
cadmium-containing solution.  Because the experiment was stopped after the passage of 100
bed volumes through  the column, it is not  possible to state explicitly what volume of solution
could be treated before cadmium breakthrough would occur. However, experience has shown
that if  a test  material  is capable of treating  at least 100 bed volumes of  metal-bearing
water,  use of that material is economically feasible. The essential point is that AlgaSORB®
removed cadmium well below those levels which  are allowed in drinking water.  The current
drinking water levels  for cadmium stand at 0.005 mg/L.

     After  100 bed volumes of the cadmium-containing solution had  passed through  the
AlgaSORB®-containing column, cadmium  was stripped from the column  by passing 0.15M
H2SO4 through the column.   Analysis of the column  effluents showed that nearly 90 percent
of the  cadmium was stripped  from the column  with the passage of two bed  volumes of
sulfuric acid Jhrough  the column.  Most of the remainder of the cadmium appeared  in  the
next two bed volumes.  Mass  balance calculations  showed that, within  experimental error,
all of the bound-cadmium was stripped from the column.

2.   Removal  of Copper  from Contaminated Groundwaters Containing  Halogenated
     Hydrocarbons

     Bio-Recovery Systems obtained groundwaters  which had been  contaminated with
copper, tetrachloroethylene  and dichloroethylene by a printed circuit board  manufacturer.
These waters contained a total dissolved solid content (TDS) of nearly 2000 ppm and had a
                                        10

-------
total calcium and magnesium content of approximately 300 ppm. Past experience had shown
that ion exchange resins  were not effective  in treating these waters for copper removal
because of i) the high mineral content and ii) the propensity of the resins to become clogged
with the organics  in these  waters.  However, experiments showed that 400 bed volumes of
the copper containing waters could be passed through  a column (0.7 cm i.d. x 13 cm high)
containing  AlgaSORB® without effluents from the column containing more than 0.01 ppm of
copper. The experiments were stopped at 400 bed volumes, so undoubtedly larger volumes
of waters  could have been treated before unacceptable levels of copper appeared in the
effluents.

     After  400 bed volumes had been passed through the AlgaSORB® column, the bound
copper was, within experimental  error,  completely stripped from the column  by passing
0.5M H2SO4 through the column.   Again,  as with  the  previously  described  cadmium
stripping, the copper was almost completely  stripped  within the first few bed  volumes of
eluent.

3.   Removal of Mercury from Contaminated Groundwaters

     Bio-Recovery was provided with water  samples  from  a  mercury-contaminated
groundwater site.   The site had been contaminated with mercury years ago as a result of a
process used to manufacture chlorine from seawater.  The groundwaters contained 2-3 ppm
of mercury  (both inorganic and organic mercury), had a  total  dissolved solid  content of
7,200 mg/L and contained  over 900  mg/L of  calcium  and magnesium.  Passage of these
mercury-containing  waters through  an AlgaSORB®  column (0.7 cm i.d.  x 13 cm high)
resulted in effluents which contained mercury at levels below 0.006 mg/L as determined by
analysis using cold vapor generation  and  atomic absorption spectrometry.  The customer
requires effluents of  below  0.01 mg/L  for discharge.

     These experiments show, as had earlier experiments, that  AlgaSORB® is effective in
removing both inorganic and organic mercury from aqueous solutions even in the presence of
very high concentrations of calcium, magnesium and other dissolved salts.

4.   Selective Removal of Lead from Wastewaters

     The  printed  circuit board  industry frequently plates a  tin-lead alloy onto printed
circuit boards as a base for solder connections.  The tin-lead alloy is plated from a solder
bath which often  contains  tin  and lead fluoborates.   Since tin discharge  is not currently
federally regulated, the  major  problem in treating rinsewaters derived from  tin-lead solder
baths is lead removal. One particular AlgaSORB® preparation is especially amenable for this
application  since it strongly binds lead  and allows the majority of the tin to pass through.

     A sample of a tin-lead plating  bath was obtained from a printed  circuit board
manufacturer. The bath composition  included lead fluoborate,  stannous fluoborate, boric
acid and peptone.  The bath rinsewaters commonly contain 10-60 mg/L  of lead  and about
twice as much tin.

     A column containing  AlgaSORB® (3.3  mL total bed volume)  was prepared and the tin-
lead containing waters (27.4 mg/L of lead; 49  mg/L of  tin) which had first been  adjusted to
pH  5.0 were passed through the  column at a flow rate of one-third of a bed volume  per
minute.  Two-bed volume fractions of the effluent were collected, and each of these fractions
was analyzed for tin and lead by atomic absorption techniques.  All effluent fractions showed
                                        11

-------
lead concentrations at  or below the detection limit of 0.1  mg/L  for the first 300 bed
volumes, after which lead began to  appear in the effluents.  Influent tin-lead passage was
stopped after passage of 325 bed volumes through the column after which the column was
stripped of lead by elution with 0.5M  nitric acid (8).

     All fractions eluted  through the AlgaSORB® column were also analyzed for tin. Because
tin is more weakly bound than lead, tin  began to exit the column after passage of only 33 bed
volumes of influent. Thus the AlgaSORB® column showed marked preference for lead over
tin. When the column was stripped of lead (after 325 bed volumes) the small amount of tin
bound on the column was also fully  recovered in the  nitric acid stripping solution (8).
                                        12

-------
    V. DESCRIPTION OF SITE CONTAINING MERCURY-CONTAMINATED GROUNDWATERS
       A number of  years  ago  an industrial  process using mercury resulted  in  soil
contamination with elemental mercury.  The mercury subsequently percolated through the
soils and contaminated groundwater.  At some point the mercury was oxidized to the bivalent
oxidation state and was found at various concentrations in the groundwaters depending upon
the monitoring site.  Currently, the groundwaters are  extracted  from an upper perched
groundwater table via a drainage gallery.  A facility has been constructed to treat extracted
groundwaters by the use of precipitation with  dithiocarbamates, followed by polishing with
activated carbon and a specialty ion exchange resin. The  water is pumped from the gallery at
mercury  concentrations  of 0.1-3.0 ppm and  is currently treated to  allowable  discharge
limits of  10  ppb mercury.

       Wells monitoring the groundwater during the late  1980's showed seasonal variations
in  the mercury concentrations.  It appears that mercury levels decrease in the dry seasons
compared to the rainy season.  Chemical speciation of the mercury in the groundwaters was
not rigorously determined, but speciation studies  on  soils  overlying the groundwater
indicated the predominant species was oxidized inorganic  mercury.  The composition of other
elements in  the groundwater seems to change with  the seasons  as  well, but an average
composition  is given in Table 1.  Variations in  mercury content over a four year  monitoring
period in waters from two wells, about 150 feet from one  another, are shown  in Table 2.
      ;
      TABLE 1.  AVERAGE COMPOSITION OF MERCURY-CONTAINING GROUNDWATERS
                 Constituent
Concentrations  (ma/L)
Chloride
Sodium
Calcium
Magnesium
Total Dissolved Solids
PH
5,800
2,900
460
440
11,000
8.0
       Several hypotheses  concerning mercury  speciation  in the  groundwaters were
considered by other contractors in the mid-1980's.  Based upon  available  groundwater
chemistry data and the presence of high chloride ion concentrations, it was considered likely
that  the predominant dissolved inorganic forms of mercury included  chloride complexes.
They were thought to vary from HgCI+ through  HgCU"2-  Uncomplexed ionic mercury could
be either divalent or monovalent.
                                        13

-------
   TABLE 2. SEASONAL VARIATION OF MERCURY CONCENTRATION IN MONITORING WELLS
Month/Yr
Oct/1
Nov/1
Dec/1
Jan/2
Mar/2
Apr/2
May/2
Sep/2
Dec/2
Feb/3
Sep/3
Dec/3
Apr/4
May/4
Jun/4
Aug/4
Sep/4
Oct/4
Well 1
(mtf/L)
9.60
3.35
0.29
5.50
3.80
10.00
4.20
7.70
6.10
6.20
8.50
2.70
4.00
4.00
4.40
5.80
7.70
13.00
Well 2
(mg/L)
0.370
0.293
0.426
0.230
0.390
0.200
0.300
0.370
0.510
0.500
0.240
0.140
-
0.260
0.170
0.180
0.086
0.240
Furthermore,  with many different anions present in the water, inorganic mercury could be
present in a variety of complexed forms.

       It was also established in the mid-1980's   that  the  groundwaters  contained
significant quantities of organic compounds.   It is therefore possible that some of the
mercury in the groundwater could also be in the form of organo-mercury complexes.  Less
than one percent of the mercury present in soils at the site was found to be organo-mercury.
However for an aggregate of several ppm in the recovered groundwater, even less than one
percent organo-mercury could be important considering the maximum allowable discharge
concentration  was 10 ppb mercury.  This was one of the reasons  that activated carbon was
selected as a part of the  treatment system.   Rather than  spend a great  deal  of time  in
determining mercury speciation  in the groundwaters,  it was decided to approach the
problem  on a direct,  empirical  basis.  This  led  to the current  waste  treatment  process
involving precipitation,  carbon adsorption and ion exchange.
                                       14

-------
                              VI. LABORATORY TESTING
A.     Experimental  Procedures

       Mercury analyses were performed using the EPA Method 245.1 of cold vapor atomic
absorption  spectroscopy  (9) with the exception that sodium borohydride was used as  a
reductant rather  than stannous sulfate,  upon  the recommendation  of  the instrument
manufacturer,  Perkin  Elmer.  The validity of this modification  in EPA  Method 245.1 was
substantiated by experiments described in Section VIII.

       A Perkin Elmer Model 3030B AAS instrument was calibrated daily for mercury, and
a calibration verification record was  maintained using data collected by the analysis of EPA
certified check standards.  Preparation of standards  for mercury analysis was  performed in
accordance with the specifications in Methods for the Chemical Analysis of Water and Wastes
(9).  Spiked samples were  analyzed with  each  batch of samples to determine if matrix
interference existed, and frequent blanks were run to ensure there was no mercury carry
over during analysis.
  i
       Mercury concentrations in  groundwaters,   column effluents  and  regenerating
solutions were determined  by  linear regression  calibration curves  generated from four
point standard  calibration analysis  (9).

       Samples collected  in the field pilot studies were split and sent to Woodward-Clyde
Consultants, EER Technologies and  Bio-Recovery Systems for mercury analysis.

       Laboratory tests  on the efficiency of mercury  adsorption on  AlgaSORB® were
conducted  using  small glass columns  (1.5  cm i.d. x 20 cm) which contained 25.0 mL of
sorbent.  Mercury-containing groundwaters were pumped  through the column  at flow rates
which varied from 6-20 bed  volumes per hour.  Effluents from  the  columns were collected
using a fraction collector and mercury content was determined.  Once the columns became
saturated or leaked mercury above discharge limits (10 ppb), the column was  stripped with
10 bed volumes of a selected stripping reagent followed by 10 bed volumes  of deionized
water.  Analyses  of stripping effluents were performed to verify stripping.

       More complete experimental  procedures and data analyses are found in Section VIII.
Quality Assurance.
                                          15

-------
B.     Results

1.     Water Analysis

       Samples of groundwater were collected at various times during 1989.   With one
exception  all samples were acidified to pH 2 with nitric acid in the field prior to transport
for  laboratory studies.  Once the samples were received at Bio-Recovery Systems, the
solutions were  neutralized  to the original or desired  pH with  dilute sodium  hydroxide.
Laboratory and field studies were complicated by the  fact that over a 10 month  period,
mercury concentrations changed  by an order of  magnitude.  Table  3  shows mercury
concentration variation  over the sampling period. While variations in mercury speciation
were  not  determined, laboratory studies with AlgaSORB® implied that  the mercury
speciation  varied over the sampling period.  (See below).

               TABLE 3. MERCURY CONCENTRATIONS IN GROUNDWATERS


Sample Number
103-13089
176-42089
177-42089-1
177-42089-2
265-070589
343-090189
368-100489
369-100489

Original
pH
8.5
8.0
8.0
8.0
7.9
7.8
7.9
7.9
Mercury
Concentration
ffifl/n
150
435
144
215
1120
620
1550
1550

Date
Collected
01-30-89
04-20-89
04-20-89
04-20-89
07-05-89
08-31-89
10-04-89
10-04-89
       Variations in mercury  content  of samples  176-42089, 177-42089-1  and  177-
42089-2  are  due to the method of preservation.  Two five-gallon water samples  were
collected on April 20, 1989.  One sample, 177-42089-1, was not acidified in the field and
was transported unpreserved to Bio-Recovery where 5  L was removed for testing.   The
remainder  of  sample  177-42089-1  was then acidified to pH  2,  stored for use,  and
designated  as sample 177-42089-2.  Sample  176-42089  was acidified in the field and was
transported to Bio-Recovery  Systems for  testing.   It  is clear that some mercury was lost
(perhaps due  to  container-wall  adsorption)   from   sample  177-42089-1.    Upon
acidification of the sample a slight increase in  the mercury concentration was observed.

       The waters  shown in Table 3  were used for subsequent  laboratory  tests  with
AlgaSORB®.  Water samples were adjusted to various pH values and reanalyzed for mercury
just prior to AlgaSORB® testing. Thus mercury concentrations shown in subsequent tables
may vary slightly from those shown in Table 3.
                                       16

-------
2.    AlgaSORB® Tests

      Acidified groundwater  samples collected on  January 30, 1989  (Sample 103-
13089) were adjusted to pH 6  and were pumped through an AlgaSORB®-602 column at a
flow rate  of 10 bed volumes per hour.  Table 4 shows mercury contents in the effluents
were well below the 10 ppb discharge limit through the passage of over 200 bed volumes of
sample.  Table 4 also shows results

   TABLE 4.  ANALYSIS OF EFFLUENTS FROM A COLUMN PACKED WITH AlgaSORB®-602*
      Bed Volume
      of Effluent
Ha fua/L)
 Spiked
Ha (u.a/L}
Recovery (%}   Error (%)
         1 -4
         5-8
         5-8t
         9-12
        13-16
        21-24
       105-108
       121-124
       141-144
       141-144t
       161-164
       181-184
       185-188
       201-204
       221-225
       241-244
       256-260
 0.6
 0.8
 7.8t
 0.5
 0.5
 0.8
 2.1
 2.7
 2.0
 7.7
 4.4
 4.6
 1.7
 3.5
11.7
30.0
16.7
 0
10.0
  70
30
 0
10.0
  57
43
*  Influent mercury concentration was 150 \ig/L at pH 6.0.
f  QA samples
                        Water sample 103-13089
for matrix spikes. Once 260 bed volumes of groundwater were passed through the column,
attempts were made to strip the column with 3.0 M sodium chloride. Table 5 shows results
of stripping experiments.  While some mercury was stripped with sodium  chloride, mass
balance calculations showed that only 30 percent of the loaded mercury was recovered in
stripping.  Based upon this poor recovery, sodium chloride was deemed to be inappropriate
as a stripping agent.
                                      17

-------
   TABLE 5. ANALYSIS OF STRIPPING EFFLUENTS FROM COLUMN LOADED IN TABLE 4*
                   Bed Volumes
                    of Effluent
                                Ha  (u.a/U
                    1 -4
                    5-8
                    9-12
                   13-16
                   17-20
                                1290
                                 515
                                 208
                                    1.
                                    0.8
*  Stripping solution was 3.0 M NaCI.

      A second column  of AlgaSORB®-602  was prepared and  groundwater  sample
103-13089 which was adjusted to pH 5 was loaded onto the column at a flow rate of 10 bed
volumes per hour.  Table 6 shows results of mercury analysis of effluent fractions.

   TABLE 6. ANALYSIS OF EFFLUENTS FROM A COLUMN PACKED WITH AlgaSORB®-602*
      Bed Volumes
      of Effluent
Ha (u.a/L\
Spiked
Ha (ua/Ll
Recovery (%)
Error (%)
         1 -4
        17-20
        37-40
        37-40t
        57-60
        73-76
        77-80
        93-96
       1 13-116
       133-136
       133-136t
       149-152
        0.50
        0.80
        0.65
       10.7t
        4.0
        2.2
        5.6
        2.3
        3.0
        2.5
        9.9t
        6.5
     0
    10.0
     0
    10.0
   100
    74
  26
*  Influent mercury concentration was 150 jxg/L at pH 5.0.
t  QA samples
                              Water sample 103-13089.
Good mercury retention by the AlgaSORB® was observed through the passage of 152 bed
volumes of groundwater.  Similar mercury  binding performance was observed  at pH  6
(Table 4) and at pH 5.0 (Table 6).
                                     18

-------
        Sample 177-42089-1 (unpreserved at pH 8.0) was adjusted  to pH 5.0  and was
loaded onto an AlgaSORB®-602 column at a flow rate of 10 bed volumes per hour.  A total of
168 bed volumes of effluent was collected and analyzed for mercury.  Table  7 shows results
of these analyses.  After passage of 168  bed volumes, mercury concentration in the effluent
was 27 ppb, which  is  a much higher leakage rate than observed with the same adsorbent on
sample  103-13089.  (Table 6 shows effluents had mercury contents below 7 ppb after
passage of 152 bed volumes of sample  103-13089.)

        Sample 176-42089 (acid preserved)  was loaded onto another AlgaSORB®-602
column at a flow rate of six bed volumes per hour and at pH 5.0.  Seventy six bed volumes of
effluent were collected, and then the column was stripped of mercury by the passage  of 10
bed volumes of 1.0 M sodium thiosulfate followed by 10 bed volumes of distilled water.  Once
the first loading and stripping cycle was completed, it was repeated  twice more.

           TABLE 7. ANALYSIS OF EFFLUENTS FROM A COLUMN PACKED WITH
                                 AlgaSORB®-602*
Bed Volumes
of Effluent
1 -4
17-20
33-36
33-36t
69-72
1 17-120
165-168
Spiked
Hg f(xg/L^ Hg (\ig/L\
42
2.0
3.8 0
14.6t 10
8.3
12.8
26.8
Recovery 1%) Error (%)



108 8



Influent mercury concentration was 144 ng/L at
QA sample
                                              5-  Water sample 177-42089.
       Table 8 shows results of mercury analysis on effluents from the three loading cycles.
Again high leakage of mercury was observed with this water sample.  Table 9 shows results
of the three  stripping cycles.  Mass balance calculations showed that 84, 88 and 76 percent
of bound mercury was stripped in  stripping cycles 1, 2, and 3, respectively.
                                       19

-------
   TABLE 8. ANALYSIS OF EFFLUENTS FROM A COLUMN PACKED WITH AlgaSORB®-602*
Cvcle
1





2





3





Bed Volumes
of Effluent
1 -4
21-24
21-24t
37-40
57-60
73-76
1 -4
21-24
21-24t
41-44
57-60
73-76
1 -4
21-24
37-40
37-40t
53-56
73-76
Ha (\ia/L)
27
22
3lt
68
88
124
23
14
23.5t
37
44
53
8.8
1 1
11.8
28t
40
68
Spiked
Hg f|ia/L> Recovery (%) Error (%)

0
10 88 12




0
10 95 5





0
10 163 63


* Influent mercury concentration was 400 jxg/L at pH 5.  Water sample 176-42089.
t QA sample

    TABLE 9. ANALYSIS OF STRIPPING EFFLUENTS FROM COLUMN LOADED IN TABLE 8
Cvcle
1




2




3




Bed Volumes
of Effluent
1 -4
5-8
9-12
13-16
17-20
1 -4
5-8
9-1 2
13-1 6
17-20
1 -4
5-8
9-1 2
13-16
17-20
Ha fp.g/L)
5380
352
171
13
2.6
5300
625
352
141
60
4730
640
278
15
1 0
                                   20

-------
        A different lot of AlgaSORB®-602 was prepared and  again tested on groundwater
sample  176-42089.  The  water  was loaded  at pH 5 onto a 25  mL  column containing
AlgaSORB®-602 and after passage of 76 bed volumes the column was stripped with 10 bed
volumes of 1.0 M sodium thiosulfate and 10 bed volumes of deionized water.  After the first
loading-stripping cycle a second loading-stripping cycle was done.  Data for loading is shown
in  Table  10  and for stripping in  Table 11.   Table  10 again shows high rates of mercury
leakage.  Stripping  of bound mercury was effective, however, with mass balance calculations
showing  that 99 and 92 percent of bound mercury were stripped in cycles  1 and  2,
respectively.

   TABLE 10. ANALYSIS OF EFFLUENTS FROM A COLUMN PACKED WITH AlgaSORB®-602*
Bed Volumes
Cvcle Of Effluent
1 1 -4
17-26
37-40
37-40f
53-56
73-76
2 1 -4
5-8
17-20
21-24
21-24f
37-40
37-40t
47-44
57-60
61 -64
69-72
73-76
Spiked
Hgf|xa/L^ Hg(|a.g/L^ Recovery (%)
9.9
10.1
6.8 0
21. 8f 10 150
14.6
31.0
77.5
1.4
3.1
2.1 0
14. 9f 10 128
7.2 0
14. 2f 10 70
8.6
7.6
10.0
7.6
11.5
Error <%}



50






28

30





* Influent mercury concentration was 400 jxg/L for Cycle 1 and 200 (ig/L for Cycle 2 and for both
  cycles the pH was 5.0.  Water sample 176-42089.
t QA samples
                                       21

-------
   TABLE 11. ANALYSIS OF STRIPPING EFFLUENTS FROM COLUMN LOADED IN TABLE 10
	 Cycle 	
1




2




Bed Volumes
	 of Effluent 	
1 -4
5-8
9-1 2
13-16
17-20
1 -4
5-8
9-1 2
13-16
17-20
Ha fua/n
6250
1020
230
16.4
5.3
2900
365
198
16.6
8.8
      AlgaSORB®-602 clearly showed different mercury binding characteristics  on water
sample 103-13089  (Table 4 and 6) as compared  to sample  176-42-89 (Table 7,  8, 10).
Unacceptable mercury leakage was observed with the 176-42089 samples  as compared to
the 103-13089.  This suggests that the mercury speciation  may have changed during the
time period between sample collections.                             >

      Different algae  have different  mercury binding  characteristics due to  different
biopolymers present in the cell walls. Thus a different AlgaSORB®, AlgaSORB®-601, was
synthesized containing a different algal  species and was  tested on  the 176-42089  waters.
Waters at pH 5.0 were loaded into an AlgaSORB®-601  column at a flow rate of  10  bed
volumes per hour.   Mercury was stripped with thiosulfate as described earlier.  Data for
four loading and  stripping cycles on AlgaSORB®-601  are shown in Tables  12 and 13.
AlgaSORB®-601 was more effective in  binding mercury than was AlgaSORB®-602.  Table
12 shows that mercury leakage was below 10 ppb during all four loading  cycles through the
passage of over 100 bed volumes of sample 176-42089.
                                       22

-------
  TABLE 12. ANALYSIS OF EFFLUENTS FROM A COLUMN PACKED WITH AlgaSORB®-601'
Cvcle
1










2










3















4







Bed Volumes of Effluent
1-4
21-24
37-40
37-40f
73-76
77-80
89-92
97-100
121-124
137-140
153-156
1-4
17-20
37-40
37-40|
17-20
68-72
73-76
85-88
101-104
117-120
132-135
1-4
21-24
21-24|
37-40
57-60
67-70
71-74
71-74f
91-94
97-100
107-110
117-120
121-124
121-124f
127-130
131-134
1-4
49-52
67-70
71-76
97-100
109-112
129-132
137-142
Ho: ((ifl/n Spike Hp (\itf/L\ Recoverv (°/0\ Error (%\
0.5
1.5
1.8 0
11. 5f 10.0 98 2
5.1
2.1
4.5
5.5
10.8
15.2
21.0
2.2
3.1
2.7 0
10.0f 9.0 82 18
3.1
8.9
3.8
5.9
9.8
16.5
31.2
0.7
1.4 0
10.2f 10.0 88 12
3.3
5.1
5.7
2.2 0
10.3 10.0 81 19
3.9
4.7
4.8
6.3
2.2 0
12.1f 10.0 00 1
4.4
4.5
1.1
5.3
7.1
2.1
3.6
5.2
7.2
7.3
   Influent mercury concentrations were 506, 502.  255 and 283 (xg/L for Cycles 1, 2, 3, 4, respectively.  All
   influents were at pH 5.0.  Water samples 176-42089 for Cycles 1 and 2;  177-42089 for Cycles 3 and 4
f  QA sample.
                                          23

-------
   TABLE 13. ANALYSIS OF STRIPPING EFFLUENTS FROM COLUMN LOADED IN TABLE 12
Cvcle
1




2




3




4




Bed Volumes
of Effluent
1 -4
5-8
9-1 2
13-16
17-20
1 -4
5-8
9-1 2
13-16
1 7-20
1 -4
5-8
9-1 2
13-16
17-20
1 -4
5-8
9-1 2
13-16
17-20
Hq (mq/L)
15,700
620
235
4
0.6
14,100
1,500
34
7.8
4.2
5,450
770
390
4.2
3.0
4,100
830
425
3.8
1.6
Mass  balance calculations showed 84, 92, 75 and 59 percent of the bound mercury was
stripped from the columns during stripping Cycles 1, 2, 3 and 4, respectively (Table 13).

       Yet a third alga was immobilized to produce AlgaSORB®-603. This adsorbent was
tested in the same manner as AlgaSORB®-602 (Tables 4, 6) and AlgaSORB®-601 (Table
12) on groundwater collected 4-20-89 as well as on a new  groundwater sample collected
7-5-89 (Sample 265-070589).  All water samples were loaded onto an AlgaSORB®-603
column at pH 5 and at flow rates of 10 bed volumes per hour.  After loading, the columns
were stripped with thiosulfate as described earlier.  Data for three loading and  stripping
cycles are shown  in Tables 14 and 15.  AlgaSORB®-603 was more  effective for mercury
removal than either AlgaSORB®-601  or AlgaSORB®-602  for Sample 176(177)-42089.
Mass  balance  calculations showed that 95,  86 and 99 percent of bound mercury was
recovered in stripping cycles 1, 2  and 3, respectively (Table 15).
                                      24

-------
  TABLE 14. ANALYSIS OF EFFLUENTS FROM A COLUMN PACKED WITH AlciaSORB®-603*
Cvcle
1
















2











3










Bed Volumes
of Effluent
1 -4
17-20
37-40
37-40f
57-60
73-76
77-80
93-96
93-96f
113-116
133-136
149-152
153-156
157-160
161-164
169-172
177-180
1 - 4
21-24
37-40
37-40f
57-60
73-76
77-80
89-92
97-100
1 17-120
137-140
149-152
1 -4
21-24
61-64
89-93
100-103
104-108
104-108t
113-116
121-124
129-132
137-140
Spike
Ha (\ig/L) Hg fp.g/L^ Recovery (%)
2.8
2.1
1.4 0
10. 8f 10.0 94
3.5
4.5
3.5
2.2 0
12.4t 10.0 102
8.0
11.7
16.6
6.2
8.1
8.0
9.9
11.1
0.5
0.9
1.0 0
8.7f 10.0 77
4.1
6.1
8.9
5.9
6.1
8.9
10.6'
14.3
6.6
1.6
3.9
8.8
10.5
4.0 0
13.2t 10.0 92
14.2
16.8
24.6
34.0
Error (%]



6




2











23














8




* Influent mercury concentration for Cycle 1 was 268 ng/L and was Sample 177-42089.  Influent
  mercury concentration  for Cycles 2 and 3 were 1160  and 910 p.g/L, respectively and  was
  Sample 265-070589.  All Cycle influents were at pH 5.0
t QA samples
                                        25

-------
   TABLE 15. ANALYSIS OF STRIPPING EFFLUENTS FROM COLUMN LOADED IN TABLE 14
Cvcle
1




2




3




Bed Volumes
of Effluent
1 -4
5-8
9-1 2
13-16
17-20
1 -4
5-8
9-1 2
13-16
17-20
1 -4
5-8
9-1 2
13-1 6
17-20
Ha (jia/U
10,800
540
192
4.4
3.8
31,000
1,250
3,200
2.0
0.8
28,200
2,290
1,250
7.0
0.6
       AlgaSORB®-602 was also tested on water Samples 265-070589.  Results of that
testing, under conditions as used for other sample testing, are shown in Table 16.

   TABLE 16. ANALYSIS OF EFFLUENTS FROM A COLUMN PACKED WITH AlgaSORB®-602*
Bed Volumes
of Effluent
1-41.3
21-24
41-44
41-44f
57-60
69-72
Ha ua/Ll

3.4
0.8
8.1f
27.0
72.5
Spike
Ha fun/Li


0
10.0


Recovery (%} Error (%)



73 27


  Influent mercury concentration was 940 ng/L at pH 5.0.  Water sample 265-070589.
t QA sample

The mercury concentration in water at the site had increased to nearly 1 mg/L by the time
sample 265-070589 was taken and AlgaSORB®-602 showed unacceptable  leakage rates.
                                      26

-------
       New water samples were collected on 9-1 -89.  Since AlgaSORB®-603 appeared to be
the best  formulation for waters collected on 4-20-89 and 7-5-89, it was tested on water
sample 343-090189.  Data are shown in Table 17.  Conditions of pH and flow rates were
those described earlier.  It  is clear from Table 17, that very high unacceptable mercury
leakage occurred.

       AlgaSORB®-603 had proved to be effective in mercury recovery from samples 177-
42089 and  265-070589 which  contained  268  ppb  and  1160  ppb,  respectively, of
mercury (Table 14).  Table 17 shows that at  mercury  levels of 620 ppb in sample 343-
090189, poor  mercury  recovery was observed with AlgaSORB®-603.  These data again
suggested that mercury speciation was changing in waters taken from the site which would
account for the variation in mercury binding for different water samples.

       Because of the inconsistency of performance of various AlgaSORB® preparations with
different water  samples, a different approach was taken.

   TABLE 17. ANALYSIS OF EFFLUENTS FROM A COLUMN PACKED WITH AlgaSORB®-603*
                      Bed Volumes
                       of Effluent
Ha (aa/Ll
                          1
                          7
                         1 3
                         1 9
    2.6
   36.0
   37.0
   42.0
* Influent mercury concentration was 620 jj.g/L at pH 5.0.  Water Sample 343-090189.

       Work  performed previous to this  study  indicated that  two other AlgaSORB®
preparations,  AlgaSORB®-624 and AlgaSORB®-640, may be effective for mercury removal
even if mercury concentration and/or mercury speciation changed in solutions.  AlgaSORB®-
624 had shown high mercury binding capacities but also rather high mercury leakage on the
order of 20-40 ppb.  AIgaSORB®-640, on the other hand, showed  rather low mercury
binding capacities, but at the same time, produced  effluents  which contained mercury in the
low ppb range. Thus two columns, one containing AlgaSORB®-624 and the other containing
AlgaSORB®-640 were prepared and connected in series.  Groundwater Sample 343-090189
was adjusted  to pH  7.9, the native pH, and was first  passed through the AlgaSORB®-624
column and then through the AlgaSORB®-640 column.  Data for these experiments are shown
in Table 18.  Table 19 shows repeat experiments of Table 18 using water sample 369-
100489, collected on October  4, 1989.
                                      27

-------
                  TABLE 18. ANALYSIS OF EFFLUENTS FROM TWO
      COLUMNS IN SERIES PACKED WITH AlgaSORB®-624 and AlgaSORB®-640*
                  Bed Volumes
                  of Final Effluent
Ha (ua/L)
12
34
43
60
80
104
113
121
130
140
159
170
180
190
200
210
230
0.0
0.0
0.6
1.8
3.3
3.4
2.9
3.9
4.4
3.2
7.5
4.0
3.5
0.1
0.1
0.1
2.3
Influent waters were sample 343-090189 (mercury concentration  620 (ig/L) for  the first 90
bed volumes.  Sample 368-100489 (mercury concentration of  1550 (xg/L)  provided influent for
bed volumes 91-230.
                                     28

-------
                  TABLE 19. ANALYSIS OF EFFLUENTS FROM
  TWO COLUMNS IN SERIES PACKED WITH AlgaSORB®-624 AND AlgaSORB®-640*
                  Bed Volumes
                                                 Hg (H9/L)
1 2
24
36
48
60
72
84
96
108
120
132
144
156
168
180
192
204
228
264
276
300
324
333
0.3
0.2
' 0.3
0.3
0.3
0.5
0.5
0.7
0.7
0.8
0.8
0.9
0.9
1.0
0.8
0.8
0.9
0.9
0.6
1.2
2.1
2.0
1 .9
Influent waters were Sample 369-100489 (mercury concentration 1550 jxg/L) at pH 7.9.
                                  29

-------
                     VII. ON-SITE, PILOT SCALE DEMONSTRATION
       On-site, pilot scale demonstrations  were conducted  using  AlgaSORB®-624  and
AlgaSORB®-640 as adsorbents.  A small portable water treatment system manufactured by
Bio-Recovery Systems was used for these studies (Figure 3).  This portable unit is designed
so that columns ranging in size from 1-4 inches in diameter can be placed on the unit.  For
the pilot testing one inch diameter columns were used. Based upon laboratory experiments
it was predicted that one-inch diameter columns would become saturated with mercury in
3-4 weeks at flow rates of 10 bed volumes per hour.

       One  column was filled with AlgaSORB®-624 and the second column was  filled with
AlgaSORB®-640.  Each column had a volume of 0.4 L.  The two columns were run in series so
that groundwater, with  no  pH adjustment, was directed first through the AlgaSORB®-624
column and  then through the AlgaSORB®-640  column.  Effluent samples were collected from
a sample port between  the two columns as well as from effluent emanating from the second
column.   Effluent samples were split into  three portions.   One  portion was sent to
Woodward-Clyde Consultants  for  immediate  mercury analysis (within 12-24  hours of
collection).   Another portion was  acid-preserved and  sent to EER Technology for mercury
analysis, while the third portion was  preserved  and sent  to  Bio-Recovery Systems for
analysis.

       On-site pilot scale testing was conducted from November 6 to December 1,  1989.
The site was available for testing only from  7:OOAM-3:30PM each day.  At the  end of a
treatment day, the system was simply shut down and then restarted the next day. Flow rates
through the system were 10 bed volumes per hour.

      By the time the  on-site testing had begun in November,  the mercury concentrations
In  the groundwaters had  changed  from  about 1500 ppb  (in  October) to 780 ppb on
November 7.   During the three week  on-site test period  the  mercury concentration
continued to vary.  Table 20 shows mercury concentration variations during the on-site test
period.  Mercury  was found to  vary from as low as 330 ppb  to as high as 1000 ppb.
                                       30

-------
Figure  3.   Portable Waste Treatment System Used for On-Site Testing

-------
                    TABLE 20. VARIATION IN MERCURY CONTENT
              OF GROUNDWATERS DURING ON-SITE PILOT SCALE TESTING
                 Pate
                       Mercury*
                  Concentration  (\ig/\-)
              11/07/89
              1 1/08/89
              11/09/89
              11/10/89
              11/14/89
              11/15/89
              1 1/16/89
              1 1/17/89
              11/20/89
              11/21/89
              1 1/27/89
              11/28/89
              11/29/89
              11/30/89
                        780
                        500
                        332
                        490
                        810
                        700
                        730
                        690
                        850
                        970
                       1000
                       1000
                        730
                        590
   Each day during on-site testing, a
   through the columns.
water sample was analyzed for mercury content before any water was pumped
      Results of mercury analyses on effluents from the complete test system, i.e., from
the effluent from the second column are shown in Table 21. Table 21 shows analytical data
for only a portion of all collected samples.  Full data with matrix spikes and QC/QA data are
found In Appendices A and B.
                                     32

-------
   TABLE 21. ON-SITE PILOT TESTING FOR MERCURY REMOVAL FROM GROUNDWATERS*
Bed Volumes
of Effluent
7-8
85-86
163-64
229-230
289-290
313-314
343-344
379-380
415-416
449-450
467-468
503-504
533-534
587-588

Bio-Reqovery
Analysis
9.5
5.3
2.1
1.4
1.8
1.9
5.5
2.0
1.8
4.9
4.0
5.8
7.7
10.5
...Mercury Concentration
Woodward-Clyde
Analysis
14.2
8.0
3.6
1.4
2.6
2.4
9.3
3.1
3.2
7.8
7.2
9.6
10.0
13.0
...fuo/U
EER Technologies
Analysis
1 1
<10
<10
<10
<10
<10
10.0
<10
<10
10.0
<10
<10
<10
15
   A portable water treatment system was equipped with two columns connected in series. The first column was filled
   with AlgaSORB®-624 and the second was filled with AlgaSORB®-640.  Groundwaters were pumped through the system at
   a flow rate of 6 bed volumes per hour.  Effluent samples were collected and sent to Woodward- Clyde Consultant, EPA
   (EER Technologies Corporation)  and Bio-Recovery systems for analysis.

       With the exception  of the first fraction collected, Table 21 shows that well over 500
bed volumes  of mercury-contaminated groundwaters  were  treated  before  mercury
concentrations in the effluents approached the 10 ppb discharge limit.

       During on-site testing, samples were collected from  the sample port between the two
columns and were sent to  Woodward-Clyde for mercury analysis. These samples represent
water treated only by AlgaSORB®-624 prior to entering the AlgaSORB®-640 column.  Data
from these analyses are shown in Table 22.  These data  show rather constant leakage of
mercury from the first  column in the range of 20-100 ppb over  the testing period.   The
data in Table 20, 21,  and 22  confirm  laboratory experiments which  showed AlgaSORB®-
624 was capable of removing the majority of the  mercury and AIgaSORB®-640 was capable
of polishing effluents from  AlgaSORB®-624 to permitted discharge levels.
                                         33

-------
         TABLE 22. ANALYSIS OF EFFLUENTS FROM THE AlgaSORB®-624
               COLUMN ON THE PORTABLE TREATMENT SYSTEM
               Bed Volumes
               of Effluent
 Mercury  Concentration*
	(ua/Ll	
                1-261
                  262
                  281
                  316
                  333
                  352
                  382
                  413
                  429
                  446
                  47,0
                  495
                  518
                  542
                  561
                  585
    Not Determined
        28
        40
        33
        38
        33
        26
        90
       120
        38
        46
        53
        54
        68
        61
       107
Analysis by Woodward-Clyde Consultants.
                                 34

-------
                             VIII. QUALITY ASSURANCE
       The objective of this program was  to demonstrate effective mercuiy removal and
recovery from groundwaters.  The  critical data needed  to support this  objective were
measurements of mercury concentrations in water prior to treatment and after treatment.
A quality assurance project plan was developed for these measurements and was approved in
December, 1988.

A.      Verification of Modification of  EPA Method 245.1 for Mercury Analysis

       Since the manufacturer of the cold vapor apparatus used in this study recommended
the use of sodium borohydride instead of stannous sulfate or stannous chloride as a reducing
agent,  initial experiments were designed to verify the validity of using sodium borohydride
as a reductant.

       Two standard stock solutions containing  mercury  at a concentration of 1000 ppm
were purchased, one from VWR and the other from J. T. Baker. The VWR standard was used
solely by the analyst while the J.T. Baker standard sample was used solely by the QA chemist
for spikes.

       In  initial tests  a  100 ppb serial dilution of the VWR mercury standard was  prepared
by the  project supervisor.  This 100  ppb sample was  used by the analyst to calibrate the
atomic  absorption spectrometer and by the QA chemist to  prepare  spiked samples to check
calibration.  These experiments were designed to  verify that techniques employed by the
analyst and QA chemist  were comparable. Results are shown in Table 23.            ,

    TABLE 23. MERCURY ANALYSIS OF STANDARDS USING SODIUM BOROHYDRIDE AS A
                                    REDUCTANT
Sample
1
2
3
4
5
6
Actual Mercury
Concentration (fig/L)
6.0
6.0
12.0
12.0
18.0
18.0
Analyzed Mercury
Concentration fjig/L}
6.0
6.0
11.3
10.6
15.4
16.1
Percent
Error
0.0
0.0
6
1 1
1 4
1 1
       A second series of experiments were designed whereby the project supervisor
prepared a 100 ppb mercury-containing sample from the VWR stock for the analyst and a
                                       35

-------
100 ppb  mercury-containing  sample from the J.T. Baker stock for the  QA chemist.  The
analyst used his 100 ppb sample to calibrate the  instrument and the QA chemist used her
sample for spikes to check calibration. Results of these experiments are shown in Table 24.

    TABLE 24. MERCURY ANALYSIS OF STANDARDS USING SODIUM BOROHYDRIDE AS A
                                    REDUCTANT
Sample
1
2
3
4
5
6
7
8
9
1 0
1 1
12
Actual Mercury
Concentration <\ia/L)
6.0
6.0
11.0
11.0
16.0
16.0
6.0
6.0
12.0
12.0
18.0
18.0
Analyzed Mercury
Concentration (]iq/L
5.7
5.6
10.7
9.5
15.3
15.8
5.1
5.3
10.0
11.3
16.7
16.7
Percent
Error
5
7
3
1 4
4
1
1 5
1 2
1 7
6
7
7
B.     Analysis of EPA-Provided Standard

       The  EPA Environmental Monitoring  Systems  Laboratory in  Cincinnati sent
Bio-Recovery Systems a standard Water Pollution Quality Control Sample for testing.  The
sample contained 15 different metal ions including mercury which.was present  both  in
inorganic and organic forms.  The ampule containing the standards was opened by snapping
the top at the break area on  the neck, and 10.0 mL of the concentrate was transferred to a
1.0 L volumetric flask, brought to volume  and analyzed.  Actual concentrations of metals in
the sample are shown in Table 25. Actual mercury content in the EPA sample was 5.0 u.g/L.
Results from Bio-Recovery analysis of the sample are shown  in Table 26.  According  to EPA,
analyzed  mercury  values  must  fall within the  range of 3.85-6.25 u,g/L in  order to be
within  the  95  percent confidence interval.  Table  26 shows that 8 of the 11  analytical
determinations for mercury were  within the 95 percent confidence level.
                                        36

-------
                  TABLE 25. ERA-PROVIDED SAMPLE INFORMATION



                        U.S. Environmental Protection Agency

              Environmental Monitoring Systems Laboratory - Cincinnati

                    WATER POLLUTION QUALITY CONTROL SAMPLE

                        TRUE VALUES FOR TRACE METALS -1

      The true values (T.V.) given below represent the actual weighing and all subsequent
dilutions as given in the sample  preparation instructions.  The mean (X), standard deviation
(S)  and  95%  confidence  interval  (X  ±2S) are  calculated  from regression  equations
generated from date from previous Performance Evaluation Studies. Table 25 represents
the  statistics  when the sample preparation instructions are followed.

                STATISTICS USING SAMPLE PREPARATION INSTRUCTION
                           (All values expressed as u,g/L)
Parameter
S
Al
As
Be
Cd
Co
Cr
Cu
Fe
Hg
Mn
Ni
Pb
Se
V
Zn
500
100
100
25
100
1 00
100
100
5.0
100
100
100
25
250
100
506.0
99.2
99.4
24.4
99.5
99.8
99.1
100.2
5.05
98.8
100.4
100.1
22.8
250.9
99.8
39.4
9.60
5.37
1.64
6.31
7.68
4.83
8.78
0.60
5.21
6.20
7.50
2.73
15.5
5.44
427
80.0 -
88.7 -
21.2 -
86.8 -
84.4 -
89.4 -
82.7 -
3.85-
88.4 -
88.0 -
85.1 -
17.4 -
220
89.0 -
585
118
1 10
27.7
1 1 2
115
109
118
6.25
109
113
115
28.3
282
1 1 1
                                      37

-------
                     TABLE 26. MERCURY ANALYSIS OF EPA
                 WATER POLLUTION QUALITY CONTROL SAMPLE*
Trial
Number
1
2
3
4
5
6
7
8
9
1 0
1 1
Analyzed Mercury
Concentration (\ig/L)
6.4
6.9
6.1
6.7
5.4
5.3
5.2
5.1
5.2
4.6
4.8
Within 95 percent
Confidence Interval
No -
No -
Yes
No -
Yes
Yes
Yes
Yes
Yes
Yes
Yes
2.4%>6.25
10.4%>6.25

7.2%>6.25







The actual mercury contraction in the sample was 5.0 (ig/L.
confidence level is 3.85-6.25
The accepted range at 95 percent
                                    38

-------
C.     Mercury Spikes

      During the course of testing various  AlgaSORB® preparations for efficiency of
mercury binding, the analyst was given samples of groundwater effluents from AlgaSORB®
columns which had been spiked by the QA chemist with amounts of mercury unknown to the
analyst.   Section VI  shows tables including the amount of spiked mercury as well as the
percent  error and  the  percent  recovery of the mercury spikes.   However Table 27.
summarizes all mercury spikes.  From a total of 36 spiked samples, analysis of 26 samples

           TABLE 27. ERROR AND RECOVERY ANALYSIS OF MERCURY SPIKES
Spike
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
, 10
10
. .. 10
10
10
10
10
10
10
10
10
10
Percent
Error
32
43
3
0.5
26
27
67
213
23
40
12
5
63
50
28
30
3
19
12
19
1
42
15
215
19
0.5
8
6
2
23
8
7
130
147
27
10
Percent
Recovery
68
57
97
100.5
74
127
167
313
123
140
84
95
163
150
128 ,
70
97
81
88
81
99
142
115
315
81
100.5
108
94
102
77
92
93
230
247
73
110
                                      39

-------
were within  the  allowable 35 percent error range giving  a 73% accuracy level on spike
recovery.

D.     Mercury Analysis in the Presence of Thiosulfate.

       During the course of stripping the bound mercury from the AlgaSORB® columns using
1.0 M  sodium thiosulfate, an  analytical problem  was encountered.   The  presence of
thiosulfate appeared to  interfere with mercury analysis (Table 28.)

              TABLE 28. EFFECT OF THIOSULFATE ON MERCURY ANALYSIS*
Actual Mercury
ffia/D
0
1000
2000
Analyzed Mercury
(\IQ/L]
1
356
528
Percent
Error

64
74
* All mercury standard samples contained 1.0 M Na2S2O3

       Further  investigation  revealed that  acid  digestion of samples containing high
concentrations  of thiosulfate  produced the  interference.  Thus  attempts were made to
alleviate the interference by oxidizing the thiosulfate with hydrogen peroxide at different
pHs prior  to acid digestion.  ' Results  of these experiments, shown in  Table  29 indicated
peroxide oxidation did not alleviate the problem.

  TABLE 29. ANALYSIS OF MERCURY-THIOSULFATE SAMPLES OXIDIZED WITH HYDROGEN
                                    PEROXIDE*
Oxidation
pH
2
5
8
2
5
8
2
5
8
2
5
Ratio of Peroxide
to Thiosulfate (Molart
1.0
1.0
1.0
2.0
2.0
2.0
5.0
5.0
5.0
10.0
10.0
Actual
Mercurv <[ia/L)
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
Analyzed
Mercurv (\ia/L)
270
155
210
240
130
105
290
150
160
105
105
Percent
Error
73
85
79
76
87
90
71
85
84
90
90
  All mercury standard samples were in presence of 1.0 M Na2S2O3
                                       40

-------
        The analytical  interference problem was finally overcome by eliminating the acid
 digestion as prescribed in EPA Method 245.1. Table 30 shows results of these analyses.

              TABLE 30. MERCURY ANALYSES OF THIOSULFATE CONTAINING
                         SOLUTIONS WITHOUT ACID DIGESTION*
1 0
20
1000
1000
1000
500
500
500
1 0
5
1000
1000
1000
500
500
500
8.2
16.2
1070
1070
1020
540
540
510
9.3
4.5
1010
1030
1060
560
520
530
	 roiocui E:nui
1 8
1 9
7
7
2
8
8
2
7
1 0 ,
1
3
6
1 2
4
6
*  All mercury standard samples contained 1.0 M Na2S2O3

       Table 30 clearly shows that elimination of the acid digestion step also eliminated the
interference in the mercury analysis.  Thus all AlgaSORB® column eluents resulting from
stripping with thiosulfate were analyzed without the acid digestion step.

E.     Analysis of Samples Resulting from On-Site Testing.

       During  on-site  pilot scale  testing of AlgaSORB®  for  mercury  recovery  from
groundwaters,  Effluents from A!gaSORB®-containing columns were collected, preserved,
split and sent to  EER Technologies (Cincinnati), Woodward-Clyde Consultants (Oakland) and
Bio-Recovery Systems for mercury analysis.  Results from  Bio-Recovery Systems analysis
and QC data have been reported earlier in Section VII.  Sample numbers, and bed volumes of
column  effluent and influent to which sample numbers correspond are listed  in Table 31.
Appendices A  and B show mercury  analysis and QC data for Woodward-Clyde  and  EER
Technologies, respectively.
                                       41

-------
                  TABLE 31. IDENTIFICATION OF SAMPLES SENT
           TO WOODWARD-CLYDE CONSULTANTS AND EER TECHNOLOGIES
                           FOR MERCURY ANALYSIS
Sample Number
436-110789
437-110789
438-110789
439-110789
440-110789
441-110789
442-110789
443-110789
444-110889
445-110889
446-110889
447-110889
448-110889
449-110889
450-110889
451-110889
452-110889
453-110889
457-110989
458-110989
459-110989
460-110989
461-110989
462-110989
463-110989
464-110989
465-110989
466-110989
Description*
Influent
Blank
1-2 BV
7-8 BV
13-14 BV
19-20 BV
25-26 BV
31-32 BV
37-38 BV
43-44 BV
49-50 BV
55-56 BV
61-62 BV
67-68 BV
73-74 BV
79-80 BV
Blank
Influent
85-86 BV
90-92 BV
97-98 BV
103-104 BV
109-110 BV
115-116 BV
121-122 BV
127-128 BV
Blank
Influent
Hq (ua/U
780
0.4
0.5
14.2
2.6
2.4
2.2
3.7
4.1
7.1
7.1
7.6
7.3
8.1
8.0
8.1
ND
500
8.0
8.4
10.4
10.7
10.4
10.4
10.9
10.5
ND
332
Samole Number
473-111389
474-111389
475-111389
476-111389
477-111389
478-111389
479-111389
480-111389
481-111389
482-111389
487-111489
488-111489
489-111489
490-111489
491-111489
492-111489
493-111489
494-111489
495-111589
496-111589
497-111589
498-111589
499-111589
500-111589
501-111589
502-111589
503-111689
504-111689
Description
Blank
Influent
133-134 BV
139-140 BV
145-146 BV
151-152 BV
157-158 BV
163-164 BV
169-170 BV
175-176 BV
Blank
Influent
181-182 BV
187-188 BV
193-194 BV
199-200 BV
205-206 BV
211.212 BV
217-218 BV
223-224 BV
229-230 BV
235-236 BV
241-242 BV
247-248 BV
Blank
Influent
253-254 BV
259-260 BV
* Hq fua/U
0.5
490
13.0
3.3
2.8
3.1
3.0
3.6
3.0
3.1
ND
810
2.5
2.7
4.8
2.5
2.2
2.7
4.1
2.3
1.4
2.1
2.3
2.7
ND
700
4.3
2.6
BV, unless otherwise indicated, designates bed
collected into a single fraction.
volumes of effluent from the second column
                                    42

-------
                                TABLE 31.  - continued
Sample Number
505-111689
506-111689
507-111689
508-111689
509-111689
510-111689
511-111689
512-111689


513-111689


514-111789
515-111789
516-111789
517-111789
518-111789
519-111789
520-111789
521-111789
522-111789
526-111789
524-111789


525-111789


526-112089


Description*
265-266 BV
271-272 BV
277-278 BV
283-284 BV
289-290 BV
Blank
Influent
Lead Col Effluent
@ 262 BV

Lead Col Effluent
@281 BV

295-296 BV
301-302 BV
307-308 BV
313-314 BV
319-320 BV
325-326 BV
331-332 BV
337-338 BV
Blank
Influent
Lead Col Effluent
@316 BV

Lead Col Effluent
@ 333 BV

Influent


Hg (\ig/L)
2.6
2.7
2.6
2.9
2.6
ND
730

28


40

4.0
2.4
2.4
2.4
2.5
2.3
2.4
2.8
ND
690

33


38

850


Sample Number Descriotion* Hn fim/l \
527-112089
528-112089
529-112089
530-112089
531-112089
532-112089
533-112089
534-112089

535-112089


536-112089

537-112089
538-112189
539-112189
540-112189
541-112189
542-112189
543-112189
544-112189
545-112189
546-112189

547-112189

548-112189


549-112189

343-344 BV
349-350 BV
355-356 BV
361-362 BV
367-368 BV
373-374 BV
379-380 BV
385-386 BV

Lead Col Effluent
@ 352 BV

Lead Col Effluent
@ 382 BV
Blank
Influent
Blank
391-392 BV
397-398 BV
403-404 BV
409-410 BV
415-416 BV
421-422 BV
427-428 BV

431-432 BV

Lead Col Effluent
@413BV

Lead Col Effluent
@ 429 BV
9.3
4.1
0.3
0.5
0.8
2.6
3.1
4.1


33


26
3.0
970
1.3
4.3
3.4
6.3
4.6
3,2
2.9
2.7

2.5


90


120
BV, unless otherwise indicated, designates  bed volumes  of  effluent from the second column
collected into a single fraction.
                                         43

-------
                               TABLE 31. - continued
Samel© Number
550-112789
551-112789
552-112789
553-112789
554-112789
555-112789

556-112889
557-112889
558-112889

559-112889

560-112889

561-112889

562-112889

563-112889

564-112889

565-112889

566-112889

567-112889

588-112989
589-112989

590-112989
Description*
Influent 1
Blank
437-438 BV
443.444 BV
449-450 BV
Lead Col Effluent
@ 446 BV
Influent 1
Blank
455-456 BV

461-462 BV

467-468 BV

473-474 BV

479-480 BV

485-486 BV

491-492 BV

497-498 BV

Lead Cot Effluent
@ 470 BV
Lead Col Effluent
@ 495 BV
Influent
Blank

503-504 BV
Ha (u.a/U
,000
0.1
12.2
8.0
7.1

38
,000
1.0
10.5

7.7

7.2

6.9

7.2

7.5

7.5

7.7


46

53
730
.08

9.6
Sample Number
591-112989
592-112989
593-112989
594-112989
595-112989
596-112989
597-112989
598-112989

599-112989


600-113089

601-113089

602-113089

603-113089

604-113089

605-1 13089

606-113089

607-113089
608-113089
609-113089

610-113089


Description
509-510 BV
515-516 BV
521-522 BV
527-528 BV
533-534 BV
539-540 BV
545-546 BV
Lead Col Effluent
@ 518 BV
Lead Col Effluent
@ 542 BV

Influent

Blank

551-552 BV

557-558 BV

563-564 BV

569-570 BV

575-576 BV

581-582 BV
587-588 BV
Lead Col Effluent
@ 561 BV
Lead Col Effluent
@ 585 BV

* Hg (MI/IT)
10.1
9.7
9.9
10.3
10.7
10.7
10.6

54

68

590

.08

13.9

12.1

12.8

13.2

13.2

13.2
13.0

61.0

107.0

BV, unless otherwise indicated, designates  bed volumes  of effluent from the second  column
collected into a single fraction.
                                         44

-------
                                   IX. REFERENCES
 1.  Hanson, B., J. Haley, C. Enfield, and J.GIass, "Effectiveness of Groundwater Extraction-
     Technical Consideration, Field  Experience,  Policy Implications, Proc. 10th National
     Conference Superfund  '89, November 27-29,  1989, Hazardous Materials Control
     Research  Institute,  Silver Spring, MD., 1989, pp. 501-502.

 2.  Darnall, D.W., B. Greene, M. Hosea, R.A.  McPherson, M. Henzl and M.D. Alexander
     "Recovery of Heavy Metal Ions  by Immobilized Alga," in   Trace Metal  Removal frnm
     Aqueous Solution,  R. Thompson ed.. London:  Royal Society  of  Chemistry, Special
     Publication No.  61, pp. 1-24 (1986).

 3.  Greene, B.  and D. W. Darnall. "Algae for Metal Binding," in Microbial  Metal Recovery.
     H. Ehrlich,  J.  Brierley, and C. Brierley, eds., New York,  NY:  McGraw-Hill, 277-302
     (1 990).

 4.  Robinson,  P.K., A.L. Mabe and M.D. Trevan, "Immobilized Algae" A Review  Process
    Biochemistry 21:  122-127 (1986).

 5. Bedell, G.W.  and  D.W.  Darnall, "Immobilization  of Non-Viable, Biosorbent Algal
    Bio-mass  for. the Recovery of Metal Ions", in Biosorbents and Blosorptlnn Recovery of
    Heavy MelalS. B. Volesky ed., Boca Raton, FL: CRC Press, in press. (1990).

 6.  Darnall, D.W., B. Greene, M. Henzl, J.M. Hosea, R.A. McPherson, J.  Sneddon and M D
    Alexander,  "Selective Recovery of Gold and Other Metal Ions  from an Algal Biomass"'
    Environmental  Science and  Ter.hnnlngy ?n-  206-208 (1986).

 7.  Greene, B.,M, Hosea, R.McPherson, M. Henzl, M.D. Alexander and  D.W.  Darnall
     Interaction of Gold (1) and Gold  (111) Complexes with Algal Biomass", Environmental
    Science and Technology 20:627-632 (1986).

8.  Darnall, D.W., A.M. Gabel and J. Gardea-Torresdey.  "AlgaSORB®  A New Biotechnology
    for Removing  and Recovering Heavy  Metal Ions from  Groundwater and Industrial
    Wastewater , in Proc. of the 1989 A & WMA/EPA  Intl. Symp. on  Hazardous Waste
    Treatment:  Biosystems  for  Pollution Control, Air  & Waste  Management Association
    Pittsburgh,  pp  113-124 (1989).

9.  Methods for the Chemical Analysis of  Water and Wastes.  EPA-600/4-79-020 US
    Environmental  Protection Agency,  Revised  March  1983  and  subsequent  EPA-600/4
    Technical Additions Thereto, Cincinnati, Ohio, 1983.
                                        45
                                             •&U.S. GOVERNMENT PRINTING OFFICE: 1990 - 748-159/20426

-------

-------

-------
Agency
Cincinnati OH 45268
                                                                                                                                                     BULK RATE
                                                                                                                                              POSTAGE  & FEES PAID
                                                                                                                                                       •  EPA
                                                                                                                                                  PERMIT No. G-35
Official  Business
Penalty for Private Use, $300
                                                               Please make all necessary changes on the above label,
                                                               detach or copy, and return to the address in the upper
                                                               left-hand corner.

                                                               If you do not wish to receive these reports CHECK HERE Q;
                                                               detach, or copy this cover, and return to the address in the
                                                               upper left-hand corner
                                                             EPA/540/5-90/005a

-------