UNITED STATES ENVIRONMENTAL PROTECTION AGENCY WASHINGTON, D.C. 20460
Aprii 12,2017
TH E A D M IN ISTR A TO R
Mr. Harry M. Johnson Hunton & Williams, LLP Riverfront Plaza, East Tower 951 East Byrd Street Richmond, Virginia 23129-4074
Mr. Major Clark Mr. Kevin Bromberg U.S. Small Business Administration Office of Advocacy 409 3rd Street, SW, 7th Floor Washington, D.C. 20416
Re: Petitions for Agency Reconsideration and Stay o f Effluent Guidelines for the Steam Electric Point Source Category
Dear Mr. Johnson, Mr. Clark and Mr. Bromberg:
This letter concerns petitions from the Utility Water Action Group dated March 24,2017, and the U.S. Small Business Administration dated April 5,2017, to the U.S. Environmental Protection Agency requesting reconsideration and an administrative stay of provisions o f the EPA's final rule titled "Effluent Limitations Guidelines and Standards for the Steam Electric Power Generating Point Source Category," 80 FR 67838 (November 3,2015).
After considering your petitions, I have decided that it is appropriate and in the public interest to reconsider the rule. The EPA is acting promptly to issue an administrative stay o f the compliance dates in the rule that have not yet passed pending judicial review, pursuant to Section 705 o f the Administrative Procedure Act. This stay will be effective upon publication in the Federal Register, The EPA also intends to request that the U.S. Court o f Appeals for the Fifth Circuit stay the pending litigation on the rule for 120 days (until September 12, 2017), by which time the agency intends to inform the Court o f the portions o f the rule, if any, that it seeks to have remanded to the agency for further rulemaking, after careful consideration o f the merits in your petitions. Also, because an administrative stay lasts only during the pendency o f judicial review, the EPA intends to conduct notice and comment rulemaking during the reconsideration period to stay or
Internet A ddress (U R L) e http://w w w .epa.gov H ec y c le d /R e cy c la tH * Printed w ith V egetable O il B ased Inks on 100% Postconsum er. Process Chlorine Free Recycled Paper
amend the compliance deadlines for the rule. This letter does not address the merits of, or suggest a concession o f error on, any issue raised in the petitions.
As part o f the reconsideration process, should the EPA conduct a rulemaking to amend the rule or any part o f it, the EPA expects to provide an opportunity for notice and comment.
If you have questions regarding the reconsideration process, please contact Sarah Greenwalt at (202) 564-1722. If you have any questions or wish to discuss the litigation, please have your counsel direct inquiries to Jessica O'Donnell at (202) 305-0851.
Respectfully yours,
E. Scott Pruitt
WHuHnUtAoMnS
April 12,2017
IIUNION & WH IJAMS U P RIVI RI RON I PI.A/A, I AST TOWhR 981 I AST BYRD STRI I I RICHMOND. VIRGINIA 23219-4074
'ITT 804 788 8200 I AX 804 788 8218
HARRY M. JOHNSON, III DIRUTI DIAU 804 788 8784 MA1I : pjohnson't/ lmnton.com
l it i NO: 2 9 142.07(012
By U.S. M ail and E-mail
Docket No. EPA-HQ-OW -2009-0819
The Honorable E. Scott Pruitt Administrator United States Environmental Protection Agency William Jefferson Clinton Building 1200 Pennsylvania Avenue, N. W. Mail Code: 1101A Washington, DC 20460
Supplemental Inform ation in Support of the Utility W ater Act G roup's Petition for Rulemaking to Reconsider and Administratively Stay the Effluent Limitations
Guidelines and Standards for the Steam Electric Power Generating Point Source Category; Final Rule, 80 Fed. Reg. 67,838 (Nov. 3, 2015)
Dear Administrator Pruitt:
Enclosed please find a study by the Electric Power Research Institute entitled Biological Treatment o f Flue Gas Desulfurization Wastewater at a Power Plant Burning Powder River Basin Coal. We are submitting this study in support of the Utility Water Act Group's Petition for Rulemaking to Reconsider and Administratively Stay the Effluent Limitations Guidelines and Standards for the Steam Electric Power Generating Point Source Category; Final Rule, 80 Fed. Reg. 67,838 (Nov. 3, 2015).
As discussed in our Petition, this study examined the performance of a biological treatment pilot system for flue gas desulfurization wastewater (FGDW) installed at a power plant burning 100% Powder River Basin (PRB) coal. See Petition, pp. 48-49. The results of the study support UWAG's contention that the model FGDW biological treatment used by EPA as a basis for limits within the ELG rule is not demonstrated for FGDW generated by plants burning subbituminous coals such as PRB. The enclosed study, to our knowledge, is the only study of biological treatment at a facility burning 100% PRB coal. The study shows
A l l A M A AUSTIN BANGKOK BITJING BRI SSUUS CUARTO ITT D A M A S HOUSTON LONDON I OS ANGUI US Mci.UAN MIAMI NI W YORK NORIOUK RAI I.IGII RICHMOND SAN FRANCISCO TOKYO WASHINGTON w\\ \v.hunton.com
Hunton WILLIAMS
The Honorable Scott Pruitt April 12. 2017 Page 2
the pilot was unsuccessful in attaining consistent compliance with the rule's monthly average selenium limit for FGDW. For most of the pilot study--despite significant adjustments to ordinary operations, including a 50/50 dilution of the FGDW prior to treatment--the pilot results indicate a failure to meet the selenium limit.
A copy of this study has also been electronically mailed to the Office of Water Docket Center for filing in Docket No. EPA-HQ-OW-2009-0819.
Additionally , we will be submitting a supplemental filing in support of our Petition tomorrow.
Please contact me if you have any questions about the study or the Petition.
Sincerelv,
Enclosure
cc by hand delivery and e-mail: Mr. Michael H. Shapiro
cc by e-mail: Jessica O'Donnell. Esq. Kevin S. Minoli, Esq. EPA Docket Center
Harry M. Johnson, 111
29142.070312 MMI US (>40680850
E R G I l l f CTRIC POWER RESEARCH INSTITUTE
Biological Treatment of Flue Gas Desulfurization Wastewater at a Power Plant Burning Powder River Basin Coal
Pilot Demonstration with the ABMet Technology
Biological Treatment of Flue Gas Desulfurization Wastewater at a Power Plant Burning Powder River Basin Coal
Pilot Demonstration with the ABMet Technology
3002006089 Final Report, March 2017
EPRI Project Manager J. Preece
ELECTRIC POWER RESEARCH INSTITUTE 3420 Hillview Avenue, Palo Alto, California 94304-1338 PO Box 10412, Palo Alto, California 94303-0813 * USA
800.313.3774 650.855 2121 * askepri@epri com www epri com
DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES
THIS DOCUMENT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN ACCOUNT OF WORK SPONSORED OR COSPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE. INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, THE ORGANIZATION(S) BELOW. NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM: (A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD. PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS DOCUMENT, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL PROPERTY, OR (III) THAT THIS DOCUMENT IS SUITABLE TO ANY PARTICULAR USER'S CIRCUMSTANCE; OR (B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING ANY CONSEQUENTIAL DAMAGES. EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS DOCUMENT OR ANY INFORMATION, APPARATUS. METHOD. PROCESS. OR SIMILAR ITEM DISCLOSED IN THIS DOCUMENT. REFERENCE HEREIN TO ANY SPECIFIC COMMERCIAL PRODUCT. PROCESS, OR SERVICE BY ITS TRADE NAME, TRADEMARK, MANUFACTURER, OR OTHERWISE, DOES NOT NECESSARILY CONSTITUTE OR IMPLY ITS ENDORSEMENT, RECOMMENDATION, OR FAVORING BY EPRI. THE FOLLOWING ORGANIZATION, UNDER CONTRACT TO EPRI. PREPARED THIS REPORT: CH2M Hill
NOTE
For further information about EPRI, call the EPRI Customer Assistance Center at 800.313.3774 or e-mail askepri@epri.com. Electric Power Research Institute, EPRI, and TOGETHER...SHAPING THE FUTURE OF ELECTRICITY are registered service marks of the Electric Power Research Institute, Inc. Copyright 2017 Electric Power Research Institute, Inc. All rights reserved.
ACKNOWLEDGMENTS
The following organizations, under contract to the Electric Power Research Institute (EPR1), prepared this report:
CH2M Hill
6600 Peachtree Dunwoody Road, Bldg. 400
Atlanta, GA 30328
*
Principal Investigators J. Siczka S. Graziano B. Bratt K. Munirathinam K. Doup
The Electric Power Research Institute (EPRI) also contributed to the preparation of this report.
Principal Investigator J. Preece
This report describes research sponsored by EPRI.
This publication is a corporate document that should be cited in the literature in the following manner:
Biological Treatment o f Flue Gas Desulfurization Wastewater at a Power Plant Burning Powder
River Basin Coal: Pilot Demonstration with the ABMet Technology. EPRI, Palo Alto, CA: 2017.
3002006089.
'
PRODUCT DESCRIPTION
This report documents a pilot demonstration conducted to evaluate treatment of wet flue gas desulfurization (FGD) wastewater at a power plant burning Powder River Basin (PRB) coal. FGD wastewater at PRB sites tends to have very high concentrations of total dissolved solids (TDS), calcium, magnesium, sulfate, nitrate, selenium, and various metals. The low chlorine content of PRB coal enables high cycles of concentration in the FGD absorber because absorber cycles are generally limited on chloride concentration due to specific metallurgy requirements. The resulting wastewater is typically challenging to treat due to the high TDS and high hardness scaling potential.
Background The U.S. Environmental Protection Agency's (EPA's) revision of effluent limitations guidelines (ELGs) in 2015 has resulted in new requirements for FGD wastewater. Under the rule, EPA established limits based on physical/chemical (P/C) and biological treatment technologies with numeric limits on total arsenic, total mercury, total selenium, and nitrate/nitrite as nitrogen. Limited data are currently available on the treatment of FGD wastewater from PRB coal plants using biological treatment technologies.
O b je c tiv e s A pilot system was operated to evaluate and study processes in treating FGD wastewater, using pilot effluent water quality targets associated with the monthly average limits from the U.S. EPA 2015 ELGs. Target concentrations for arsenic, mercury, selenium, and nitrate/nitrite as nitrogen from the pilot system treated effluent were established as 50% of the 2015 ELG monthly average limits to account for initial blending with service water. Analytical and field data were collected to study performance of biological treatment systems on FGD wastewater at a PRB coal site.
Approach For this demonstration, GE Water & Process Technologies (GE) provided and operated a pilot treatment system consisting of a membrane bioreactor (MBR) process, for nitrification and denitrification, and the Advanced Biological Metals Removal (ABMet) system for denitrification and selenium reduction. The pilot systems were evaluated using a small continuous portion of effluent from the host plant's full-scale P/C treatment system. After initial evaluations of the P/C effluent, GE determined the scaling potential was too high and would create issues in the small pilot equipment. GE also determined that the TDS was too high to support efficient bacteria growth and performance for reducing ammonia, nitrates, and selenium to the demonstration targets. The result of the initial ev aluation was a decision to blend the P/C treatment effluent with plant service water (untreated water from Lake Michigan) in equal proportion, and send the combined stream to the pilot system.
v
Results The ABMet pilot was initially operated with a total empty bed contact time (EBCT) of 18.8 hours vs. GE's commercial system target of 8 hours. Upon achieving acclimation, GE increased the flow rate through the treatment system to more closely simulate full-scale conditions. After the initial change, the total EBCT for the ABMet pilot was 12.4 hours and remained unchanged due to performance degradation that followed. The pilot system met the effluent targets for the first 3.5 weeks but performance quickly degraded after the flow rate was increased. It is suspected that inhibitory compounds affecting bacteria performance may be present in the FGD wastewater; however, despite various attempts to reduce inhibitory compounds, no resolution was achieved. The ABMet system produced effluent selenium concentrations three times higher than the 6 pg/L target (between 15 and 21 pg/E total selenium). A combination of factors may have resulted in the loss of selenium reduction performance, including:
Oxidants - Data are limited on specific oxidation compounds present in FGD wastewater and how they may impact biological treatment. Surrogate data on ORP, COD, and BOD were monitored, but do not provide sufficient clarity or causality.
Non-biodegradable organics - Based on data collected during the test (COD and nitrogen compound concentrations), it is plausible that organic species were present that are potentially inhibitory to the ABMet bacteria.
Nitrogen species removal (ammonia, nitrate, nitrite, and organic nitrogen) - This was the first known test using MBR for pretreatment of FGD wastewater and the first known test using a separate denitrification step before ABMet.
Sulfate-reducing bacteria (SRB) - SRB are known to exist at ABMet conditions due to low ORP conditions and are evident by the presence of sulfide production. It is plausible that the high concentration of sulfates in the site's FGD wastewater, along with the nutrients used, encouraged SRB to suppress selenium-reducing bacteria growth.
Analytical issues with total and dissolved selenium measurements were identified during the project. In some samples, values for total selenium reported low, giving false indications of water quality. Suggested studies to evaluate analytical interferences are presented. Following data validation, the analytical results show that the pilot did not meet the effluent targets for 13 weeks, and the ABMet system was never operated at full-scale conditions.
Applications, Value, and Use The methods used in this study may provide value for similar biological treatment technology evaluations. Due to site-specific conditions, each site should conduct its own evaluations.
Keywords Flue gas desulfurization Wastewater treatment Selenium removal ABMet Effluent Limitation Guidelines Biological treatment
vi
E R R I ELECTRIC POWER RESEARCH INSTITUTE
EXECUTIVE SUMMARY
Deliverable Number: 3002006089 Product Type: Technical Report
Product Title: Biological Treatm ent of Flue Gas Desulfurization W astew ater at a Power Plant Burning Powder River Basin Coal: Pilot Dem onstration with the ABMet T e c h n o lo g y *
PRIMARY AUDIENCE: Power plant operators, chemists, and engineers and environmental managers
KEY RESEARCH QUESTION
The main objectives of the pilot system were to evaluate and study processes in treating FGD wastewater, using pilot effluent water quality targets associated with the monthly average limits from the U S Environmental Protection Agency (EPA) 2015 effluent limitations guidelines (ELGs). This project investigated selenium removal using EPA's selected best available technology (BAT) in order to help plants consistently meet ELG limits, and specifically for plants burning Powder River Basin (PRB) coals.
RESEARCH OVERVIEW
Effluent from the host site's full-scale physical/chemical treatment system was collected and blended with service water (originating from Lake Michigan and untreated) at a 1:1 ratio and supplied to a pilot system designed and operated by GE Water & Process Technologies. Prior to beginning the pilot evaluation. GE determined that dilution of the full-scale physical/chemical treatment system effluent is required in order to minimize the potential negative impacts of scaling, high TDS concentrations, and toxicity in the processes The pilot system consisted of a membrane biological reactor (MBR) process with two anoxic zones, an aerobic zone with GE s ZeeWeed 500D membranes, and a final anoxic stage The MBR process was designed mainly to reduce nitrate/nitrite concentrations prior to the Advanced Biological Metals Removal (ABMet) process. The pilot ABMet system consisted of two attached growth packed-bed biofilters in series, each containing granular activated carbon (GAC) media. Laboratory analyses were conducted on many sample streams in the pilot process, and data were analyzed with process operating data collected throughout the pilot evaluation.
KEY FINDINGS
The pilot test did not consistently demonstrate that the MBR/ABMet combination of technologies adequately treats selenium to meet the 2015 ELGs limits.
100% dilution of FGD wastewater from the effluent of the host site's full-scale physical/chemical treatment system was required for the MBR/ABMet pilot system
Complete denitrification (reduction of nitrate nitrogen to nitrogen gas) was achievable in the pretreatment MBR
Selenium, present mainly in dissolved form, was significantly reduced by the pretreatment MBR (approximately 75% on average for the entire demonstration)
Analytical issues with total and dissolved selenium measurements were identified during the project. Following data validation, the analytical results show that the pilot did not meet the effluent targets for
13 weeks, and the ABMet system was never operated at full-scale conditions.
VII
ELECTRIC POWER
OE R i a i RESEARCH INSTITUTE
EXECUTIVE SUMMARY
WHY THIS MATTERS
This study provided the opportunity for EPRI and its members to evaluate selenium removal in an especially challenging FGD wastewater at a power plant burning PRB coal The results, learnings, and analytical data from the study will continue to support ongoing research related to biological treatment of FGD wastewater.
HOW TO APPLY RESULTS
Due to site-specific conditions that do not translate to every application, each site should conduct its own evaluations related to FGD wastewater treatment. The methods used in this project may provide guidance for users to apply similar approaches when evaluating biological treatment technologies.
LEARNING AND ENGAGEMENT OPPORTUNITIES Related projects are
Wet Flue Gas Desulfurization Wastewater Physical/Chemical Treatment Guidelines. EPRI. Palo Alto. CA: 2016. 3002008515.
EPRI Technical Manual: Guidance for Assessing Wastewater Impacts o f FGD Scrubbers. EPRI, Palo Alto. CA: 2006. 1013313.
Flue Gas Desulfurization (FGD) Wastewater Characterization and Management: Preliminary Summary. 2007 Update. EPRI. Palo Alto. CA. 2008. 1014073
Case Studies to Evaluate FGD Wastewater Physical/Chemical Treatment Performance EPRI. Palo Alto CA: 2013. 3002001202
Flue Gas Desulfurization (FGD) Wastewater Chemical Precipitation Treatment Performance Characterization Study. EPRI. Palo Alto. CA. 2015. 3002006251.
Pilot Evaluation of the ABMct Technology for Flue Gas Desulphurization Wastewater Treatment EPRI. Palo Alto. CA: 2014. 3002004550.
EPRI CONTACTS: Jeffery Preece. Technical Leader, jpreece@epri.com
PROGRAM: Water Management Technology, P185
Together. ..Shaping the Future of Electricity
Electric Power Research Institute 3420 Hillview Avenue Palo Alto. California 94304-1338 PO Box 10412 Palo Alto California 94303-0813 USA
800 313 3774 650 855 2121 askepn@epri com www epri com
2017 Electric Power ResearcMnslitute (EPRb Inc All rights reserved Electnc Power Research Institute EPRI and TOGETHER SHAPING THE FUTURE OF ELECTRICITY are registered service marks of the Electric Power Research Institute Inc
ACRONYMS AND ABBREVIATIONS
c F < Mg/L Hin pS/cm BOD COD DO DOC EBCT ELG EPA EPRI FGD GAC GPD GPM HRT Inf/EQ MBR mg/L mg-N/L MLSS mS/cm mV
degrees Celsius degrees Fahrenheit less than or fewer than microgram per liter micrometer microSiemen per centimeter biochemical oxygen demand chemical oxygen demand dissolved oxygen dissolved organic carbon empty bed contact time Effluent Limitations Guidelines U.S. Environmental Protection Agency Electric Power Research Institute, Inc. flue gas desulfurization granular activated carbon gallons per day gallons per minute hydraulic retention time influent equalization membrane biological reactor milligram per liter milligram of nitrogen per liter mixed liquor suspended solids microSiemen per centimeter millivolt
IX
N NA ng/L NO;,/NO:-N O&G ORP PAC P4 Phys/Chem RAS Se Se(IV) Se(VI) SeCN SeMe SRT S.U. TCLP TDS TKN TMP TOC TSS WQBEL ZVI
nitrogen not applicable nanogram per liter nitrate/nitrite nitrogen oil and grease oxidation reduction potential powdered activated carbon Pleasant Prairie Power Plant physical/chemical return activated sludge selenium selenite selenate selenocyanate selenomethionine solids retention time standard unit toxicity characteristic leaching procedure total dissolved solids total Kjeldahl nitrogen transmembrane pressure total organic carbon total suspended solids water quality-based effluent limit zero-valent iron
X
CONTENTS
1 1NTRODUCTION.................................................................................................................... 1-1 Project Understanding................................................................................................................. 1-1 Pilot Objectives............................................................................................................................. 1-2
2 BACKGROUND.......................................................................................................................2-1 P4 Existing Full-Scale Wastewater Treatment Plant Description............................................2-1 Treating Wet Flue Gas Desulfurization W astewater................................................................ 2-3 Summary of Treatment Systems and Equipment......................................................................2-3 ZeeWeed 500D Membrane Biological Reactor................................................................... 2-3 A B M e t.......................................................................................................................................2-6
3 PILOT DEMONSTRATION DESIGN......................................................................................3-1 Pilot System Design.................................................................................................................... 3-1 Equalization T an k................................................................................................................... 3-3 Pretreatment Membrane Biological Reactor.........................................................................3-3 A B M e t....................................................................................................................................... 3-6 Pilot Schedule and Conditions.................................................................................................... 3-8 Field and Laboratory Analytical Plans........................................................................................ 3-9
4 RESULTS AND DISCUSSION................................................................................................4-1 Pilot Influent Water Quality......................................................................................................... 4-1 Membrane Biological Reactor..................................................................................................... 4-3 Operating Parameters.............................................................................................................4-3 Treated Water Characteristics...............................................................................................4-6 Nitrate/Nitrite.................................................................................................................... 4-11 Selenium...........................................................................................................................4-12 Mercury.............................................................................................................................4-13 Arsenic..............................................................................................................................4-14 ABM et.........................................................................................................................................4-14
xi
Operating Parameters...........................................................................................................4-14 Treated Water Characteristics.............................................................................................4-18
Nitrate/Nitrite.................................................................................................................... 4-21 S e le n iu m ........................................................................................................................... 4-23 Mercury............................................................................................................................. 4-27 Arsenic.............................................................................................................................. 4-28 5 SUMMARY AND CONCLUSIONS.......................................................................................... 5-1 Pilot Summary.............................................................................................................................. 5-1 MBR Conclusions.........................................................................................................................5-4 ABMet Conclusions..................................................................................................................... 5-4 Recommendations for Additional Research.............................................................................. 5-6 6 REFERENCES........................................................................................................................6-9 A SAMPLING AND LABORATORY ANALYTICAL PLAN...................................................... A-1 B ANALTYICAL QA/QC DISCUSSION.................................................................................... B-1 CRAW DATA........................................................................................................................... C-1
xii
LIST OF FIGURES
Figure 2-1 P4 FGD wastewater treatment plant process flow diagram....................................... 2-2 Figure 2-2 ZeeWeed membrane fiber and ZeeWeed 500 module...............................................2-4 Figure 2-3 ABMet biofilter cross-section......................................................................................... 2-6 Figure 2-4 ABMet selenium reducing bacteria with selenium particles....................................... 2-7 Figure 3-1 Pilot process block flow diagram...................................................................................3-2 Figure 3-2 ZeeWeed 500 MBR pilot................................................................................................3-6 Figure 3-3 ABMet pilot equipment................................................................................................... 3-8 Figure 4-1 MBR influent and effluent COD and phosphorus........................................................ 4-9 Figure 4-2 MBR influent and effluent ammonia and T K N ...........................................................4-10 Figure 4-3 MBR influent and effluent nitrate/ntrite as nitrogen.................................................. 4-12 Figure 4-4 MBR influent and effluent selenium............................................................................ 4-13 Figure 4-5 MBR influent and effluent mercury............................................................................. 4-14 Figure 4-6 ABMet inline ORP trends............................................................................................ 4-16 Figure 4-7 ABMet headloss trends................................................................................................4-18 Figure 4-8 ABMet influent and effluent COD and phosphorus................................................... 4-21 Figure 4-9 ABMet influent and effluent ammonia and TKN.........................................................4-22 Figure 4-10 ABMet influent and effluent nitrate/nitrite as nitrogen............................................4-22 Figure 4-11 ABMet influent and effluent selenium......................................................................4-23 Figure 4-12 ABMet effluent sulfide and selenium .......................................................................4-25 Figure 4-13 ABMet effluent selenium.......................................................................................... 4-26 Figure 4-14 ABMet influent and effluent m ercury.......................................................................4-27 Figure 4-15 ABMet influent and effluent arsenic.........................................................................4-28
xiii
LIST OF TABLES
Table 1-1 ELG Limits on FGD W astewater................................................................................... 1-1 Table 1-2 Effluent Water Quality Objectives.................................................................................. 1-3 Table 3-1 Pilot MBR System Description........................................................................................3-4 Table 3-2 Pilot ABMet System Description.................................................................................... 3-6 Table 3-3 Field Analyses.................................................................................................................. 3-9 Table 3-4 We Energies Lab Analyses..........................................................................................3-10 Table 3-5 TestAmerica Lab Analyses...........................................................................................3-11 Table 3-6 Brooks Applied Lab Analyses.......................................................................................3-11 Table 4-1 Blended Pilot Influent Water Quality..............................................................................4-2 Table 4-2 MBR Operational Parameters........................................................................................4-3 Table 4-3 MBR Treated Effluent (Permeate) Water Quality1 .......................................................4-7 Table 4-4 ABMet Operational Parameters................................................................................... 4-15 Table 4-5 ABMet Nutrient Dosages.............................................................................................. 4-17 Table 4-6 ABMet Backwash Analysis..........................................................................................4-18 Table 4-7 ABMet Stage 2 Effluent Water Quality1...................................................................... 4-19 Table 4-8 Comparison of ABMet Effluent Selenium Concentrations During
Period of Matrix Spike Recovery Failures............................................................................4-24 Table 5-1 Summary of pilot results................................................................................................. 5-3 Table A-1 Analytical Plan - Frequency of Samples.................................................................... A-2 Table B-1 ABMet Second Stage Reactor Effluent Selenium Analysis Using Hydroxide
Preparation Method..................................................................................................................B-2 Table B-2 ABMet Effluent Selenium Speciation Results..............................................................B-4
xv
7 INTRODUCTION
Project Understanding
The U.S. Environmental Protection Agency (EPA) sets technology-based standards, termed effluent limitations guidelines (ELGs), by industry category based on various levels of technology performance, the most stringent of which is the Best Available Technology Economically Achievable (BAT). EPA revised the ELGs for the steam electric industry in 2015 to include BAT limits for flue gas desulfurization (FGD) wastewater.
The BAT limits set by EPA for FGD wastewater are based on physical/chemical and biological treatment technologies with numeric limits on total arsenic, total mercury, total selenium, and nitrate/nitrites as N. The ELGs also include Best Practicable Control Technology (BPT) limits for total suspended solids and oil & grease. EPA's ELG limits for FGD wastewater are shown in Table l - l . Additionally, water quality-based effluent limits (WQBELs) in some states are necessitating removal of selenium and other trace metals from FGD wastewater.
Table 1-1 ELG Limits on FGD Wastewater
Parameter
Units of Measure
Monthly
Daily Maximum
Average Limit
Limit
A rsenic, total
mq/ l
8
M ercury, total
ng/L
356
S elenium , total
pg/L
12
N itrate/N itrite as N
m g/L
4.4
Total S uspended S olids (TSS)
m g/L
30
Oil and G rease
m g/L
15
11 788 23 17.0 100 20
ug/L = microgram per liter mg/L = milligram per liter ng/L = nanogram per liter
The primary objective of this project was to investigate and study selenium removal using EPA's selected BAT to help plants consistently meet ELG limits, specifically for plants burning Powder River Basin (PRB) coals. The ELGs also include mercury, arsenic, and nitrate/nitrite limits; studying their removal was a secondary project goal.
Introduction
This study provided the opportunity for the Electric Power Research Institute, Inc. (EPRI) and its members to help evaluate selenium removal in an especially challenging FGD wastewater at We Energies Pleasant Prairie Power Plant (P4). The main properties of this wastewater that make treatment challenging are scaling tendency due to high sulfate and calcium concentrations; concentrations of dissolved/ionic species such as halogens, which are much higher than other FGD wastewaters from sites burning non-PRB coals; and nitrate/'nitrite-nitrogen and total Kjeldahl nitrogen1(TKN) concentrations, both of which are also much higher than other FGD wastewaters. Flence, pretreatment for reducing suspended solids, dilution of dissolved solids and a nitrification/denitrification step were included ahead of the selenium removal unit for the P4 pilot, adding to the operational complexity of the pilot system. P4 bums 100% subbituminous coal from the Powder River Basin, and has a low FGD blowdown rate for its power capacity (40 GPM), allowing the concentrations of ammonia, nitrates/nitrites, sulfates, arsenic, mercury, and selenium to reach very high levels. To EPRFs knowledge, this was the Erst time that biological treatment has been attempted on a wastewater stream from a plant that burns 100% subbituminous coal.
Pilot O bjectives The main objectives of the pilot system were to evaluate and study processes in treating FGD wastewater, with pilot effluent water quality targets equal to those listed in Table 1-2, and to validate the operating parameters for the membrane biological reactor (MBR) nitrification/denitrification pretreatment step and the Advanced Biological Metals Removal (ABMet) systems. Given the many potential changes and variations in chemistry that are inherent to FGD wastewater, the study attempted to maintain consistent selenium removal performance. When the performance changed, conditions were documented and process changes were made to recover performance.
1Ollier than uptake for cell synthesis, a selenium treatment system does not remove TKN. Therefore, if a discharge limit for ammonia nitrogen or total nitrogen is included in the permit, a TKN removal step will be required. Because ammonia removal was part of the pilot effluent water quality objectives, the reduction/removal of TKN was monitored.
1-2
Introduction
Table 1-2 Effluent Water Quality Objectives
Parameter
S elenium 1 M ercury1 A rsenic1 N itrate/N itrite N itro g e n 1 (N O 3/N O 2 -N) Total S uspended S olids (TSS) Oil and G rease (O & G ) A m m onia P h o s p h o ru s
Effluent Target2
<12 pg/L <356 ng/L
<8 pg/L <4.4 m g/L <30 m g/L <15 m g/L 1.9 to 30 m g/L <0.6 m g/L
Pilot Effluent Target3
<6 pg/L <178 ng/L
<4 pg/L <2.2 m g/L <15 m g/L <7.5 m g/L 0.45 to 15 m g/L <0.3 m g/L
Notes: 1Regulated pollutants (2015 HLG Final Rule) 2 2015 ELGs BAT Monthly Average Limits 150% of the ELG BAT Monthly Average Limits to account for dilution to reduce high TDS and scale potential < = less than/fewer than pg/L = microgram per liter mg/L = milligram per liter ng/L - nanogram per liter
1-3
2
BACKGROUND
Selenium removal from wastewater is particularly challenging because of its complicated chemistry. Selenium removal typically requires multiple treatment processes, and the species of selenium can vary throughout the different unit processes that target selenium removal. Other constituents in the wastewater may also interfere with the selenium removal processes. As FGD wastewater is a highly complex wastewater (high concentrations of many ions), there are many potential constituents that can affect treatment. Examples include oxvanions like sulfate, halogens like chloride, and nutrients like nitrogen species. Additionally, variations in coal sources, coal composition, plant operations, scrubber operations and other air emission control technologies, and other factors can cause the wastewater composition to vary. Even after successful treatment of selenium, treatment system residuals and byproducts containing selenium must be managed and disposed of properly. This section describes the treatment technologies used for the pilot demonstration.
P4 Existing Full-Scale W astew ater Treatm ent Plant Description
The full-scale P4 wastewater treatment plant was designed to treat 75 GPM of wastewater with a total suspended solids (TSS) content of less than 1.5%. An overview- of the existing physical/chemical system is provided in Figure 2-1. After equalization, wastewater flows to Reaction Tank 1 where pH is adjusted from approximately 5.6 to 8.5 S.U. by adding hydrated lime to the wastewater. Effluent from Reaction Tank 1 flows by gravity to a primary clarifier where polymer is added to aid in solids removal. Primary effluent flows to Reaction Tank 2 where METCLEAR MR2405 (commonly referred to as an organosulfide) is added for mercury precipitation. Effluent from Reaction Tank 2 flows by gravity to Reaction Tank 3 where ferric chloride is added for coagulation and for iron co-precipitation. Effluent from Reaction Tank 3 passes to a flash mix tank where polymer is added, and then to a flocculation tank, followed by a parallel plate clarifier (secondary clarifier). The plant was equipped with an upflow sand filter, but its use has been discontinued due to problems with scaling of sand, recurring fouling issues in the filters, and because it had limited effectiveness in removing mercury below the level achieved by secondary clarification alone. Optional cartridge filters can be installed after the final effluent tanks, when necessary, to filter to 0.45 microns for removal of suspended particles which may contain large amounts of mercury. Sludge is pressed in plate and frame filter presses and combined with coal combustion residuals (CCR) before being landfilled on plant property.
The average operating flow rate for the system is currently 40 GPM. Operations staff run the primary clarifier continuously when the treatment plant is in operation. The secondary clarifier mainly runs throughout the five-day work week and is shut down on weekends for cleaning (to remove scale). The two effluent discharge tanks hold approximately 10,000 gallons each and are used for storage during the weekend cleaning activities.
2-1
Background
Wastftwater Forwarding
Pump
SLUDGE HOLDING TANK
Figure 2-1 P4 FGD wastewater treatment plant process flow diagram Courtesy of We Energies
Background
Treating W et Flue G as D esulfurization W astew ater
Some technologies have been implemented full-scale at FGD wastewater treatment facilities in the United States to remove particulate (generally >0.45 pm) and/or dissolved/soluble selenium from the wastewater stream. Additional processes have been tested on a pilot scale on FGD or other selenium-containing streams. These technologies fall into two main categories:
Physical/chemical treatment technologies, in which various configurations are possible:
- Solids settling (is tank-based or in a settling pond; generally removes some particulate selenium, but is likely not sufficient to meet new ELG limits as much of the selenium tends to be in the soluble fraction)
- Zero valent iron (ZVI) contact (uses complex redox and adsorption reactions to remove soluble forms of selenium)
- Filtration (removes particulate selenium using media or ultrafiltration membranes; reverse osmosis membrane filtration can potentially remove all forms of selenium, including soluble forms)
Biological treatment:
- Fixed-film anaerobic/anoxic (where there are low' levels of dissolved oxygen [DO]) reactor-based biological treatment; generally removes various selenium species, including selenate and selenite
- Passive treatment by surface-tlow aerobic wetlands and anaerobic subsurface, vertical flow wetlands (which are generally similar in function to fixed-film systems) have been demonstrated to remove various selenium species
- Suspended growth anaerobic/anoxic systems (have recently been pilot tested in membrane biological reactors [MBRs] and moving bed biofilm reactors [MBBRs] to remove various selenium species using mechanisms similar to the established fixed-film systems)
The pilot system evaluated by EPR1 at P4 is based on active biological treatment for selenium removal, which included anoxic/aerobic MBR for nitrogen pretreatment and the GE ABMet primarily for selenium removal, tested on P4's full-scale chemical precipitation system effluent.
Sum m ary of Treatm ent System s and Equipm ent
The P4 pilot included the following treatment systems:
ZeeWeed 500D MBR Pilot
ABMet Pilot
ZeeWeed 500D Membrane Biological Reactor
In anoxic/anaerobic biological treatment, nitrate must be removed from the wastewater in order to create conditions conducive to the biological reduction of selenate and selenite to insoluble elemental selenium. Nitrate removal occurs via denitrification where nitrate/nitrite nitrogen is
Background
ultimately reduced to nitrogen gas, hut at ORP conditions higher than where reduction of senate and selenite occurs. Unfortunately, remov ing high concentrations o f nitrate using downflow reactors, such as those in ABMet. leads to plugging caused hy high bacterial solids growth and the need for more frequent backwashes. Also, because ABMet operates at low oxidation reduction potential (ORP), some ammonia-nitrogen may be formed when high nitrate is present in the influent. Additional reactors may be required to achieve the conditions needed to facilitate effective selenium reduction when high nitrate levels are present in the influent: however, site-specific evaluations must be conducted to determine the exact process suitable for treatment. Therefore, nitrate removal is key to successfully treating selenium in FGD wastewater using any of the established treatment technologies. In this P4 wastewater, the TKN concentration is also high requiring ammonia-nitrogen oxidation to nitrate-nitrogen (nitrification). I herefore. an MBR was included in the pilot system as a pretreatment nitrification/denitrification step to remove both the nitrates in the influent and nitrates generated from the nitrification process. An MBR is a combination of suspended growth activated sludge and membranes, with the latter performing the liquid/solids separation function. The membranes provide a positive barrier resulting in penneate with negligible TSS concentration. The ZeeWeed 500 module is designed to reject particles greater than 0.04 microns. Treated water is drawn through membrane by vacuum, enters the hollow fibers of the membrane, and is then pumped out of membrane. The redox conditions in stages within the activated sludge system can he manipulated to achieve target reactions (for example, carbon oxidation, nitrification, and denitrification). Target reactions can also convert dissolved species into insoluble constituents and result in rejection by the membranes. \ diagram of the ZeeWeed membrane fiber and ZeeWeed 500 module is provided in Figure 2-2.
Figure 2-2 ZeeWeed membrane fiber and ZeeWeed 500 module Courtesy of GE Water & Process Technologies
2-4
Background
Membrane permeability is maintained through multiple approaches. During normal operation, continuous air scouring from the bottom of the module is used to remove accumulated solids, containing biological activity and biomass, on the membrane surface. Mixed liquor in the membrane tank is recirculated back to the biological reactors to maintain solids distribution throughout the system. Solids are wasted from the membrane tank based on the targeted solids retention time (SRT). Treatment of the wasted solids is generally a separate process that includes dewatering and disposal of the sludge. Maintenance of the membranes includes backwashing, maintenance cleanings, and recovery clean-in-place (C1P) activities, all of which can be implemented on a set frequency or based on transmembrane pressure (TMP - calculated by the difference between the membrane influent and effluent pressures). Because the ZeeWeed 500 system is designed to maintain a constant flux, an increase in TMP indicates membrane fouling. Detailed descriptions of each cleaning option are as follows: A backwash is performed by reversing the flow through the membrane (also called "back
pulse") for a short time period using permeate (clean water produced by the membrane during normal operation) from the inside of the membrane to the outside where the suspended solids are held. For this study, aeration was also used during the backwashes. A maintenance cleaning using aeration and chemical cleaning solutions to remove more challenging material from the membrane surface that can lead to fouling. Cleaning solutions may include sodium hypochlorite, citric acid, and other traditional products used to restore membranes. These chemicals are generally recirculated, in separate batches, through the membrane system for a length of time (on the order of 15 minutes) depending on conditions. The recovery CIP process is generally conducted only one or two times per year. The process is more aggressive than maintenance cleanings with higher concentrations of chemicals and longer contact times for each batch. Defining the recovery CIP process requires an understanding of treatment and membrane conditions.
2-5
liu i 'kground
ABMet
I he ABM et system is an attached growth downflow carbon filter. It was developed bv Applied Biosciences and is currently marketed by (1: Water & Process Technologies. The ABMet system uses microbes cultured in bioreactor granular activated carbon (G A C ) beds that create anoxic/anaerobic conditions for selenate and selenite reduction to insoluble elemental selenium particles. The GAC material acts as a physical media for fixed biofilm development. A diagram o f the ABVlet biofilter is presented in Figure 2-3.
S - ____ -___
-v V V
h5- reactor cover h6- headspace Fh - biom atrix expansion volum e
h 2 - ABMet Biom atrix
Figure 2-3 ABMet biofilter cross-section Courtesy of GE Water & Process Technologies
hi - distribution systems (separate treated w ater collection and backwash supply header systems)
2-6
Background
The ABMet process uses certain heterotrophic facultative bacteria, which, in addition to reducing nitrate to nitrogen gas, reduce selenate and selenite to elemental selenium. Elemental selenium then exists as insoluble solids along with bacterial solids in the filter bed. An image of the ABMet selenium reducing bacteria and elemental selenium particles is presented in Figure 2-4.
300 nm
Figure 2-4 ABMet selenium reducing bacteria with selenium particles Courtesy of GE Water & Process Technologies
The microbes are selected as a site-specific inoculum for the bioreactors. These bacteria are mesophilic bacteria requiring the wastewater temperature to be typically between 50 degrees Fahrenheit (F) and 100F (10 degrees Celsius [C] and 38C) and pH between 6 and 7 S.U. These microbes use oxygen (O2 ) and the oxygen in the nitrate (NOT), selenate (SeO-t2'), and selenite (SeO r ) oxyanions as an electron acceptor for their respiration and an organic carbon (electron donor) source to meet their food/energv requirements. Biochemical reactions for the reduction of nitrate/nitrite and selenate/selenite (excluding biomass production) are presented below.
NOj + organic carbon -> N02 + organic carbon -* N2 + C02 + H20 2SeO\~ + organic carbon -> 2SeOf ~ + organic carbon -> SeO + C02 + H20 The carbon source can be either organic compounds present in the wastewater, typically expressed as chemical oxygen demand (COD), or an organic compound from an external source (such as methanol, acetic acid, ethanol, molasses, or corn syrup). The ABMet system typically uses a molasses-based nutrient supplement (supplied by GE), which also is a source of phosphate for the bacteria, however acetic acid was used for this study as a carbon source for both nitrate and selenium reduction processes. Nutrient addition can also impact pH of the system, where too much nutrient may cause an increase in production of organic acids and CO2 (both produced by bacteria) that lower the pH.
2-7
Background
Biological reduction converts the selenate or selenite ions into elemental selenium, which exists as a particle and is enmeshed and attached to the biofilm within the reactor. The reactors are backwashed periodically to prevent excess biomass and solids buildup and resulting hydraulic head loss, and the waste stream is then treated by dewatering the solids and disposing of the sludge with concentrated selenium. The system is also routinely degassed (similar to backwash, but for a shorter time period) to remove accumulated gases (such as carbon dioxide and nitrogen) from the filter bed to reduce hydraulic head loss. The effluent water from the ABMet unit may be aerated in a polishing aeration tank to increase the DO and to aerobically treat any excess soluble degradable COD-contributing organics, if required. Monitoring of pH and ORP is performed to control the chemical dosing and ensure selenate and selenite reduction. ORP is an indicator of biological activity in the ABMet system, where lower ORP correlates to anaerobic conditions created by biological activity in reducing oxygen, nitrate/nitrite, and selenium. Once nitrate and nitrite are fully utilized as electron acceptors, selenate, selenite, and sulfate will be used as the primary electron acceptor with remaining biodegradable COD by a different population of heterotrophic microbes. These sulfate-reducing microbe species are generally naturally occurring and their presence/biological activity results in the formation of hydrogen sulfide2.The sulfate-reducing bacteria may overgrow and outcompete selenate and selenite reducing bacteria for electron donors at high sulfate concentrations. The electron donor (organic carbon) for the ABMet system must also be carefully monitored and controlled to avoid providing too much carbon source for the sulfur-reducing microorganisms. Allowable total sulfide concentrations in the ABMet effluent should be defined during pilot testing for specific applications to minimize competition from sulfate-reducing bacteria. Solids generated from backwashing (which contain selenium) can generally be recycled upstream to the physical/chemical treatment system (if used) for removal or be treated separately. The resulting dewatered solids are typically tested using the toxicity characteristic leaching procedure (TCLP) to verify nonhazardous classification for disposal, but standards and requirements should be referenced for each site. Solids can generally be processed using solids dewatering equipment such as a plate-and-frame filter press with other solids.
2 Hydrogen sulfide is an odorous gas that is detectable at low concentrations (around 1 part per million [ppm]) and can be a human health issue at sufficient concentrations (the U.S. Department of Labor maintains exposure limits under the Occupational Safety and Health Administration standards).
3
PILOT DEMONSTRATION DESIGN
Pilot System Design
Effluent from P4's full-scale physical/chemical treatment system (that is, hydrated lime addition, primary solids removal, trace metals precipitation, and secondary solids removal) was collected and blended with service water (originating from Lake Michigan and untreated) at a 1:1 ratio in the 500-gallon pilot equalization tank. GE designed and operated the pilot system and determined that dilution of the full-scale physical/chemical effluent to minimize the potential impacts of scaling, high TDS concentration, and toxicity was required. For purposes of ELG compliance, however, this sort of dilution would require application of more stringent (i.e., lower) ELGs for constituents. Therefore, treatment without dilution would be preferable to the extent it is feasible. Blended wastewater from the equalization tank provided influent to the nitrification/denitrification pretreatment membrane biological reactor (MBR) system. Effluent from the pretreatment MBR system was collected in a feed tank for the two-stage ABMet system.
An overview of the P4 Pilot system is prov ided on Figure 3-1.
M o l Demonstration Design
Effluent from
Service
P4 FGD
Legend
-- Continuous Flow Path -- * Intermittent Flow Path
Figure 3-1 Pilot process block flow diagram
-- * Continuous Chemical Addition - Non-routine Chemical Addition
Absorber <(^)> Sample Location
Process Boundary
Pilot Demonstration Design
Equalization Tank The 500-gallon equalization tank was provided for blending o f P4's full-scale physical/chemical effluent and service water from Lake Michigan (untreated) at a 1:1 volume ratio. A combination of hand valves, flow meters, and flow control valves was used to balance the flows and achieve the targeted 1:1 volume ratio. The blended water from the equalization tank was pumped via submersible pump to anoxic zone #1 in the pretreatment MBR system. While initial provisions were made to feed different types of chemicals, only antisealant (MDC776) and potassium sulfite were added to the EQ Tank during testing. Antisealant (MDC776) - To mitigate scaling of calcium, sulfate, phosphate, and other
precipitating species. A dosage of 5 mg/L was maintained throughout the study. Potassium sulfite - For removal of potential oxidants for mitigating negative impacts to
biological activity that may inhibit selenium reduction. Addition began on January 5, 2016, near the end of Condition 2 through Condition 3. Pretreatment Membrane Biological Reactor The pretreatment MBR pilot system was seeded with 2,500 gallons of municipal mixed liquor from the Kenosha Wastewater Treatment Plant located in Kenosha, WI. The seeding occurred on October 20, 2015 and the acclimation period, after seeding, lasted four weeks. The system consisted of four zones (anoxic #1, anoxic #2, aerobic, and anoxic #3) followed by the membrane tank, which are described in Table 3-1.
JO
Pilot Demonstration Design
Table 3-1 Pilot MBR System Description
Item
Volume (gallons)
Chemical Addition
Description
A noxic z o n e #1
100 [w orking volum e]
P hosphoric acid - m acronutrient for ba cte ria ; be g a n in C o n d itio n 2 d u e to sw itch in a c e tic acid s o u rc e fro m a blended product (w hich contains phosphorus, but the concentration w as deem ed too high) to an 80% product (20% b a la n c e is w a te r). T a rg e t phosphorus concentration w as <3 m g/L in M B R p e rm e a te a nd 0.5 m g /L in the A B M e t effluent.
R eceived blended w a te r from equalization tank and m ixed liquor recirculation from the m em brane tank
M ixing provided via m echanical m ixer
A noxic zone #2
100 [w orking volum e]
None
R e c e iv e d w a te r fro m a n o x ic z o n e #1 M ixing provided via m echanical m ixer
A erobic zone
A noxic zone #3
ZeeW eed M em brane tank
1,000 [w orking volum e]
Ferric chloride - To reduce phosphorus that had been added by the acetic acid blended product, used o n ly durin g 2 -d a y period in C o n d itio n 3; originally intended for soluble arsenic precipitation if it had been needed (but soluble arsenic w as not present at concentrations higher than the target value). The blended acetic acid product contained excess phosphorus w hich m a y have led to the precipitation o f m aterial that created issues with e x c e s s s o lid s b u ild in g in th e A B M e t biofilter during testing. S caling and excess solids w ere ultim ately controlled by adjusting the p h o sp h o ro u s c o n te n t in th e nutrient.
S odium hydroxide - to increase pH
R eceived w a te r from anoxic zone #2
C ontinuous aeration provided by a blow er and fine bubble diffusers
1,000 [w orking volum e]
A cetic acid - carbon source for denitrification and partial selenate/ selenite reduction; blended type used durin g C o n d itio n 1; c h a n g e to 80% a c e tic acid m a d e in C o n d itio n 2. S w itched m ade due to high c o n c e n tra tio n o f p h o s p h o ru s in blended product.
R eceived w a te r from aerobic zone M ixing provided via m ixing pum p
270
S odium hypochlorite - m em brane
c le a n in g
C itric acid - m em brane cleaning
H ydrochloric acid - to decrease pH (b e g in n in g in C o n d itio n 2)
T hree ZeeW eed m odules provided, but o n ly o n e w a s placed in se rvice
3-4
Pilot Demonstration Design
The blended water was pumped from the equalization tank to anoxic zone #1, which had an underflow to anoxic zone #2. The water then overflowed to the aerobic tank. The aerobic tank was fitted with 10 fine bubble diffusers to achieve the desired dissolved oxygen (DO) concentration of 2 to 4 mg/L. Also, pH monitoring, with a control set point of 6 - 7 1S.U., was provided in the aerobic tank. Level controls within the aerobic tank controlled the flow to anoxic zone #3. A mixing pump and eductors were provided in anoxic zone #3 for mixing and allowed for the water to be pumped to the membrane tank. The membrane tank consisted of three 370 ft" ZeeWeed 500 membranes, which is a scaled-down version of a full-scale membrane system. Only one membrane, located in the middle module, was in service for this pilot; the other two membranes/modules were plugged and not placed into serv ice. This configuration allowed for normal operation of the air diffuser system that is located inside the membrane tank. Suction drives the treated water through the hollow libers of the membrane using an outside-in configuration. Membranes were cleaned with constant aitscouring along with periodic chemical cleans. For a full-scale system, all waste streams would be evaluated for treatment, potentially by recycle to the physical/chemical process upstream. Overflow from the membrane tank was recirculated to anoxic tank #1. The main objectives of the pretreatment MBR were as follows: Nitrify influent TKN/ammonia nitrogen - oxidation of ammonia nitrogen to nitrate nitrogen Denitrify influent and produce nitrate nitrogen - reduce nitrate/nitrite nitrogen to nitrogen gas Produce an effluent with very low TSS suitable for treatment in the downstream ABMet
system Remove insoluble selenium in influent and produced in the MBR anoxic tank Permeate from the pretreatment MBR was collected in the ABMet system's feed tank. A picture of the MBR pilot equipment is shown in Figure 3-2.
3The pilot MBR pH control set point of 6 - 7 S.U. was established by GR. C'H2M suggested that a slightly higher control range of 6.8 - 7.5 S.U. would minimize impacts on nitrification. Actual operating range, as discussed in Section 4, was 6.8 - 7.2 S.U.
3-5
P ilo t Demonstration Design
Figure 3-2 ZeeWeed 500 MBR pilot Courtesy of GE Water & Process Technologies
ABMet
The ABMet pilot system consisted o f two reactor stages (each with 1.270 I ofGAC), feed and inter-stage tanks, a backwash system, and chemical feed systems, which are described in Table 3-2. The ABMet reactors were seeded with ( i t 's bacteria on November 20, 201 5. and placed into forward flow on November 23.
Table 3-2 Pilot ABMet System Description
Item
Volume (gallons)
Chemical Addition
Description
A BM et feed tank
H ydrochloric acid to decrease pH (b e g in n in g in C o n d itio n 2)
S oda ash to increase alkalinity (C ondition 2 only)
A ntiscalant, M D C 776 (B egan at C ondition 2)
R eceived pretreatm ent MBR perm eate
ABMet stage 1
335
A cetic acid - carbon source for
denitrification and selenate/ selenite
reduction; used blended version that
contained nutrients (am m onium and
p h o s p h o ru s ) s ta rtin g in C o n d itio n 2
R eceived pretreatm ent M BR perm eate from the A B M et feed tank
S odium nitrate - nitrate (electron acceptor) source to enhance denitrification, but added only m id-w ay through C ondition 4
ABM et inter stage tank
S oda ash - to increase alkalinity (C ondition 2 only)
R eceived A B M et stage 1 effluent
3-6
Pilot Demonstration Design
Table 3-2 (continued) Pilot ABMet System Description
Item
Volume (gallons)
Chemical Addition
Description
ABM et stage 2
ABM et effluent tank
335
A cetic acid - carbon source for
R eceived A B M et stage 1 effluent
denitrification and selenate/selenite
from the A B M et inter-stage tank
reduction; used blended product
(containing acetic acid, am m onium , and
phosphorus) during C ondition 1 and 80%
acetic acid product (20% w ater) during
C on d itio n s 2, 3, and 4.
None
R eceived A B M e t stage 2 effluent
Backwash supply tank
3,000
None
R eceived A B M e t stage 2 effluent from the A B M e t effluent tank to degas and backw ash the A BM et filte rs
Backwash w aste tank
800
None
R eceived backw ash w aste from A B M et stage 1 and A B M et stage 2
Permeate from the pretreatment MBR was collected in the ABMet feed tank from where it was pumped to the top of the ABMet stage 1 reactor. Effluent from the ABMet stage 1 was collected in an inter-stage tank from where it was pumped to the top of the ABMet stage 2 reactor. Effluent from the ABMet stage 2 was collected in an effluent tank and then pumped to the backwash supply tank that allowed the effluent to overflow from the system to the effluent where all waste was returned to the P4 FGD Absorber.
The GAC media in each ABMet stage reactor provided surface area for bacteria to grow and develop a biofilm. Nutrient addition, pH control, and alkalinity adjustment were provided for each ABMet stage to provide a suitable/optimized environment for selenium reduction.
The main objectives of the ABMet system operation were as follows:
Denitrify any remaining influent nitrate nitrogen and reduce nitrate/nitrite nitrogen to nitrogen gas.
Reduce soluble selenium species (selenate and selenite) to particulate, elemental selenium and remove it from the wastewater.
P ilot Demonstration Design
Figure 3-3 ABMet pilot equipment Pilot Schedule and Conditions
( )perating data for the MBR and ABMet systems have been di\ ided into the following four conditions: Condition 1 Startup and Acclimation (October 26. 2015 to December 16,2015)- period
where bacteria are growing and becoming established in the treatment system. Condition 2 Initial Operation (December 17, 2015 to January 10,2016) operational
period defined by GF.'s original test plan. Condition 3 Increased Flou/Reduced HRT (January 11,2016 to February 13. 2016)
operational period intended to more closely simulate full-scale HR 17FBC I conditions. Condition 4 Process Adjustments (February 14, 2016 to March 3, 2016) period where GE
responded to decreased selenium removal performance of the ABMet. Increasing the flow rate under Condition 3 was a conscious decision made by the project team in an effort to simulate full-scale conditions (i.e., to match HRT/EBC'T for each process technology ). Due to factors that resulted in poor selenium removal performance, of which increased flow rate is not believed to be a sole-contributor, additional process changes were made during Condition 4 in attempt to return the pilot systems back to original selenium removal performance (which had been meeting the effluent/treatment targets). 3-8
Pilot Demonstration Design
The project team was able to extend the duration of the pilot demonstration past its original end date, but testing concluded when the P4 units entered an outage and were not able to supply FGD wastewater. For two weeks prior to pilot shutting down, the ABMet system was operated by feeding blended physical/chemical treatment effluent from the storage tanks with ABMet effluent; the MBR did not operate during this time.
Details on the process conditions, adjustments, and performance changes are provided in the following sections.
Field and Laboratory A nalytical Plans
Field and laboratory analyses were conducted to control system operation and optimize performance. Field analyses were conducted by GE throughout the demonstration on the MBR and ABMet pilot systems. A summary of field parameters and frequency of the analyses are provided in Table 3-3.
Table 3-3 Field Analyses
Parameter
C hem ical tank levels (inches)
Minimum Frequency (per week)
3
pH (S .U .) by handheld m eter
5
D issolved oxygen (m g/L) by handheld m eter
5
C onductivity (pS /cm ) by handheld m eter
5
O R P (m V ) by inline m eter
5
N itrate/nitrate (m g o f N /L) by H AC H test kit
5
Tem perature (F)
5
R aw W a te r T otalizer (gallons)
5
S ervice W ater T otalizer (gallons)
5
A n o xic zone #1 level (inches)
5
A noxic zone #2 level (inches)
5
A noxic zone #3 level (inches)
5
Laboratory analyses were conducted by We Energies' corporate lab, TestAmerica Laboratories, Inc., and Brooks Applied Labs for analytical chemistry evaluations. Parameters analyzed by each lab are summarized in Table 3-4, Table 3-5, and Table 3-6. Turnaround times for parameter results varied with each lab, but were generally one to two weeks with the exception of selenium by We Energies which was within three days. Samples were shipped to We Energies' lab and TestAmerica same-day and all samples requiring dissolved parameter analysis were filtered by the labs. Samples sent to Brooks Applied were on 24-hour delivery shipping. Samples analyzed
3-9
Pilot Demonstration Design
by We Energies and TestAmerica were taken every Monday, Tuesday, and Thursday during Conditions 1 and 2 and only Tuesday and Thursday during Condition 3. The entire analytical plan, which identifies the sample locations and frequencies, is provided in Appendix A.
Table 3-4 We Energies Lab Analyses
Parameter
Analytical Method
A lum inum , total and filtered
EPA 200.7
A rsenic, total and filtered
EPA 200.8 D R C -0 2
Boron, total and filtered
EPA 200.7
C adm ium , total and filtered C alcium , total and filtered
EPA 200.7 /E P A 200.8 EPA 200.7
C obalt, total and filtered C opper, total and filtered Iron, total and filtered
EPA 200.7 EPA 2 0 0 .7 /E P A 200.8 EPA 200.7
Lead, total and filtered M agnesium , total and filtered
EPA 2 0 0 .7 /E P A 200.8 EPA 200.7
M anganese, total and filtered
EPA 200.7
M ercury low -level, total and filtered M ercury low -level, 0.45 pm M olybdenum , total and filtered M olybdenum , filtered
EPA 1631E EPA 1631E EPA 200.7 EPA 200.7
N ickel, total and filtered
EPA 200.7
S elenium , total and filtered Silver, total and filtered
EPA 200.8 D R C -N H 3 EPA 200.7
S odium , total and filtered Thallium , total and filtered Tin, total and filtered
EPA 200.7 EPA 200.8 EPA 200.7
Zinc, total and filtered
EPA 1638 DRC
3-10
Table 3-5 TestAmerica Lab Analyses
Parameter
Analytical Method
A lkalinity A m m onia B icarbonate BOD C arbonate C hlorides, filtered COD C onductivity, pS/cm
SM2320B SM4500 SM2320B SM5210B SM2320B SM4500 SM5220C SM2510B
F luoride, total N itrate/N itrate (N) P hosphorus, total S ulfate, filtered S ulfide TDS TKN TOC TSS
SM4500 SM4500 SM4500 SM4500 SM4500 SM2540C SM4500 SM5310C SM2540D
Table 3-6 Brooks Applied Lab Analyses
Parameter
A rsenic, total and filtered A rse n ic S pciation, filte re d 1 S elenium , total and filtered S elenium S pciation, filte re d 1
Analytical Method
EPA 200.8/6060 EPA 200.8/6060 EPA 200.8/6060 EPA 200.8/6060
'Spciation analyses only conducted on pilot influent and effluent samples
4
RESULTS AND DISCUSSION
This section presents the results associated with the P4 pilot study. Raw data from the pilot stud) are presented in Appendix C. During the pilot study, influent water quality/chemistry was considered typical for the site (no major upsets/excursions or abnormal conditions occurred in any process upstream of the pilot system). The P4 wastewater treatment system was offline for two weeks in mid-November 2015 (during the MBR and ABMet acclimation period in Condition 1) due to mechanical issues, but stored effluent water provided consistent flow and chemistry to the pilot. In mid-February 2016, the P4 system shut down again through the beginning of March, but stored water (with consistent chemistry ) was provided to the pilot. The demonstration ended March 3, 2016. There were many performance changes in the pilot MBR and ABMet processes throughout the demonstration. These changes are documented and described to the best of the project team's capabilities.
Pilot Influent W ater Q uality Table 4-1 presents a summary of the blended pilot influent water quality (50 percent full-scale physical/chemical treatment effluent and 50 percent service water from Lake Michigan [untreated] by volume). The blended pilot influent total selenium concentration averaged 1,175 jug/L with a range of 770 to 2,500 pg/L. The blended pilot influent dissolved selenium concentration averaged 1,093 pg/L. Comparisons between each data set revealed most of the blended influent selenium was in the soluble form. Nitrate/nitrite nitrogen averaged 98 milligrams of nitrogen per liter (mg-N/L), with a range of 65 to 160 mg-N/L. Ammonia averaged 52 mg/L with a range of 10 to 180 mg/L4. The ranges indicate variability in the full-scale physical/chemical effluent, which was observed over the course of the pilot study. The variability could be explained in part by the site's normal maintenance process of shutting dow n the physical/chemical system every week for manual cleaning of the components. During this time, the concentrations of components build in the absorbers and a new "batch" of water is delivered to the system every Monday morning.
4The higher concentrations of ammonia may be attributed to selective catalytic reduction (SCR) layers approaching end of life at P4. causing higher ammonia injection rates to be used for meeting NO, removal requirements which subsequently increases the ammonia concentration in the FGD wastewater. SCR catalyst life decays naturally over many years until replacement is needed. The replacement is generally scheduled with a steam generator boiler outage.
4-1
Results and Discussion
Table 4-1 Blended Pilot Influent Water Quality
Parameter
Units
Total S elenium D issolved S elenium N O 2 /N O 3 -N Total M ercury D issolved M ercury Total A rsenic D issolved A rsenic pH1 Total C alcium D issolved C alcium Total M agnesium D issolved M agnesium A m m onia C onductivity COD BOD TOC C hloride F lu o rid e S ulfate O rtho-P hosphate Total P hosphorus T otal K jeldahl N itrogen (T K N ) S ulfide TSS TDS Total A lkalinity B icarbonate A lkalinity C arbonate A lkalinity
pg/L
mq/ l m g/L ng/L ng/L pg/L
ng/L S.U. m g/L m g/L m g/L m g/L m g/L p S /c m m g/L m g/L m g/L m g/L m g/L m g/L m g/L m g/L m g/L m g/L m g/L m g/L m g/L m g/L m g/L
Pilot Influent - All Operating Conditions
Minimum
Maximum
Average
770
2 ,5 0 0
1,175
710
1,900
1,093
65
160
98
20
240
72
6.9
29
13.5
0.45
8.10
4.67
0.45
8.50
4.68
6.6
7.7
400
830
479
400
840
471
1,600
3,100
2 ,0 3 9
1,600
3,100
2,023
10
180
52
11,000
2 2 ,0 0 0
13,378
360
1,600
596
1.00
2.60
1.8
5.1
77.0
28.1
2 ,3 0 0
3 ,2 0 0
2,690
18.0
43.0
24.0
6 ,2 0 0
11,000
7,805
<0.027
0.74
0.10
0.027
0.750
0.093
42
220
101
<0.22
1.70
0.15
4
31
12
12,000
2 5 ,0 0 0
16,452
78
120
103
78
120
103
<3.9
<3.9
<3.9
Notes: Blended Influent: 50 percent full-scale physical chemical treatment effluent and 50 percent service water from Lake Michigan (untreated) 'Analysis performed onsite BOD - biochemical oxygen demand pS'cm =microSiemen per centimeter S.U. = standard units TDS = total dissolved solids TOC - total organic carbon
4-2
Results and Discussion
Variability (the difference between minimum and maximum concentration values) was also observed in arsenic and mercury throughout the study. The data indicate pilot feed water contained arsenic mostly in the soluble form whereas mercury was mostly in the particulate form. The TDS ranged from 12,000 to 25,000 mg/L during the pilot study.
The following sections discuss the pilot study results, including operational parameters and effluent water quality for each operating condition.
M em brane Biological Reactor
Operating Parameters
MBR operational parameters are discussed in this section. As previously mentioned, the P4 fullscale physical/chemical treatment effluent was blended with service water from Lake Michigan at a 1:1 ratio in the pilot's equalization tank. This blended stream served as influent to the pilot system and was pumped to the MBR anoxic zone # 1.
After the initial seeding w ith biomass from the local municipal wastewater treatment plant, the mixed liquor was recirculated through the MBR for two days with 300 gallons of blended pilot feed water. The system was placed in forward flow through the MBR after these two days and acclimation continued for four weeks. Table 4-2 compares the MBR operational parameters at the four test conditions. The flow rate to the MBR averaged approximately 1 GPM during Conditions 1 and 2 and 1.5 to 1.6 GPM in Conditions 3 and 4. This resulted in combined HRT in the bioreactor zones decreasing from approximately 35 hours during Conditions 1 and 2 to approximately 23 hours during Conditions 3 and 4. The increase in flow rate and subsequent reduction in HRT was part of the initial pilot test plan. Operating at the lower flow rate initially was intended to allow for adequate acclimation of the biomass and the plan was to increase the flow rate after acclimation in order to simulate full-scale conditions.
Table 4-2 MBR Operational Parameters
Condition Description
Condition Number Parameter/Dates
Influent Flow Rate A v e ra g e Range
M BR pH C ontrol S et P oint M B R H R T A n o xic Z o n e #1 M BR H R T A noxic Z one #2 M BR HRT A erobic Zone M BR HRT A noxic Z one #3 Z e e W e e d 500 Flux
Z eeW eed 500 B ackw ash Flux M BR A eration Rate
Startup and Acclimation
1 10/26/15 to
12/16/15
1.05 G PM 1 to 2.6 GPM
6.8 to 7.2 1.6 hours 1.6 hours 15.9 hours 15.9 hours 10 GFD (first 3 w eeks) 5.8 GFD 20 GFD 12 SCFM
Initial Operation
2 12/17/15 to
1/10/16
Increased Flow/ Reduced HRT 3 1/11/16 to 2/13/16
1.01 G P M 1 to 1.25 G PM
6.8 to 7.2 1.7 hours 1.7 hours 16.5 hours 16.5 hours 5.8 G FD
1.54 G PM 1.25 to 1.6 G PM
6.8 to 7.2 1.1 h o u rs 1.1 h o u rs 10.8 hours 10.8 hours 5.8 G FD
Process Adjustments
4 2/14/16 to
3/3/16
1.6 G PM 1.6 G PM 6 .8 to 7.2 1.0 hour 1.0 hour 10.4 hours 10.4 hours 5.8 GFD
20 GFD 12 SC FM
20 GFD 12 SCFM
20 GFD 12 SCFM
4-3
Results and Discussion
Table 4-2 (continued) MBR Operational Parameters
Condition Description
Condition Number Parameter/Dates
MBR MLSS A ve ra g e Range
R AS Flow Rate S ludge W asting R ate
Startup and Acclimation
1 10/26/15 to
12/16/15
5,760 mg/L 3,400 to
6,500 m g/L 6 to 8 GPM
none
Initial Operation
2 12/17/15 to
1/10/16
3,950 m g/L 3,700 to
4,200 m g/L 6 to 8 G PM
none
Increased Flow/ Reduced HRT
3 1/11/16 to
2/13/16
2,690 m g/L 1,000 to
6,500 m g/L 6 to 8 G PM
80 GPD
Process Adjustments
4 2/14/16 to
3/3/16
4,183 mg/L 2,900 to
5,800 mg/L 6 to 8 G PM
none
M B R D O A n o xic Z one #1 A ve ra g e Range
0.93 m g/L 0.11 to 2.4 m g/L
2.55 m g/L 1.74 to 3.90
m g/L
3.37 m g/L 1.12 to 6.23 m g/L
0.12 m g/L 0.02 to 0.49 m g/L
M BR DO A noxic Zone #2 A ve ra g e Range
0.23 m g/L 0.01 to 2.06 m g/L
0.29 m g/L 0 to 2.49 m g/L
1.34 m g/L 0 to 5.53 m g/L
0.02 m g/L 0 to 0.2 m g/L
M BR DO A erobic Zone A ve ra g e Range
3.20 m g/L 0.86 to 5.90 m g/L
3.19 m g/L 1.7 to 4.93
m g/L
4.02 m g/L 0.17 to 9.00 m g/L
2.84 m g/L 0.43 to 5.22 m g/L
M BR DO A noxic Z one #3 A v e ra g e Range
0.14 m g/L 0 to 0.56 m g/L
0.02 m g/L 0 to 0.05 m g/L
0.02 m g/L 0 to 0.13 m g/L
0.01 m g/L 0 to 0.04 m g/L
M BR O R P A noxic Z one #3 A v e ra g e Range
N u m b e r o f M e m b ra n e s in S e rv ic e
-387 m V -143 to -489 m V
1
-423 m V -376 to -477 m V
1
-331 m V -58 to -427 m V
1
-354 m V -83 to -416 m V
1
C hem ical Doses H ydrochloric A cid (to Z ee W e e d T ank) Sodium H ydroxide (to A e ro b ic Z one)
950 m g/L 350 m g/L
950 m g/L 350 m g/L
400 m g/L 350 m g/L
400 m g/L none
C O D A d ded via A c e tic A cid to A n o x ic Z o n e #3
Average Added
897 m g/L
Range
563 to 1,464 m g/L
Ferric C hloride (to A erobic Z one) P h o sp h o ric A cid (to A n o xic Z one #1 )
60 m g/L (2 d a ys o n ly)
none
618 m g/L 452 to 807
m g/L none
5 m g/L
447 m g/L 227 to 565 m g/L
none
5 m g/L
643 mg/L 630 to 756 m g/L
none
5 m g/L
P otassium Sulfite (to E qualization T ank)
none
none
25 to 200 m g/L
none
Notes: GFD = gallons per square feet per day
SCFM = standard cubic feet per minute
GPM gallons per minute mV ~ millivolt
GPD gallons per day
RAS = return activated sludge (the How from Zee Weed tank to MBR Anoxic Zone )
4-4
Results and Discussion
On average, the DO was maintained above 2.8 mg/L (2-4 mg/L target) in the aerobic zone and was less than 0.1 mg/L (<0.5 mg/L target) in anoxic zone #3 during Conditions 2, 3, and 4. The ORP in anoxic zone #3 averaged less than -330 mV during all four conditions. These values were within GE's target range.
As noted in Table 3-1, the acetic acid product used to supply nutrients for the biological process was changed during the study. The blended acetic acid product was first used during Condition 1, and contains an organic carbon source (acetic acid), ammonium, and phosphorus (the latter two are nutrients). Due to the higher than anticipated phosphorus concentration (20 times higher) in the blended acetic acid, the carbon source was sw itched to an 80 percent solution of unblended acetic acid (the balance 20 percent is water) for Conditions 2, 3, and 4 to limit excess phosphorus in the final effluent. The excess phosphorous led to precipitation of excess solids which, if left untreated, could cause prolific scaling in the process equipment. To prevent scaling with the blended nutrient product, ferric chloride was added to the aerobic zone for two days until the nutrient product was changed. Afterward the acetic acid source was changed, ferric chloride dosing was stopped. Phosphoric acid was fed after the acetic acid product was changed to achieve less than 3 mg/L in the MBR permeate. The reduction in COD added by the nutrient from Conditions 1 to 2 to 3 (noted in Table 4-2) coincides with the project team's attempts to increase nitrate/nitrite concentration in the MBR permeate in attempt to improve ABMel selenium reduction. Additional discussion is provided in the ABMet section.
The ZeeWeed membrane flux was initially set to 10 GFD with an operating cycle time" of 10 minutes and resting mode duration of 30 seconds. During the first three weeks of operation in the acclimation period, the ZeeWeed TMP averaged less than 1 PSI. It was also during this time that the acetic acid feed point was inadvertent ly placed at the aerobic zone instead of the anoxic zone #3. Once the feed point was moved, an increase in TMP was observed and the membrane flux set point was reduced to 5.8 GFD. A second spike in TMP occurred gradually from December 12 through 24, 2015. Once the TMP reached 8 PSI. the system was placed in automatic standby and from December 24 through 28, the MBR and ABMet systems were offline due to the high MBR TMP. A recovery clean was completed on December 27 and 28; the recovery clean included soaking the membranes in a 1,000 mg/L solution of sodium hypochlorite, followed by soaking in a 2,000 mg/L solution of citric acid. The system was placed back into operation on December 28 and the maintenance cleaning schedule was adjusted for a citric acid cleaning every other day.
The MBR process accumulates solids that come from biomass production and naturally by precipitation of various chemical species in the wastewater. The MBR solids, referred to as mixed liquor suspended solids (MLSS), were purged from the system beginning in early January 2016 at 80 GPD. The MLSS purge rate was set to 20 gallons every 6 hours to avoid reducing flow to the ABMet. Discharge of MLSS ended in mid-February due to low MLSS concentrations and decreased selenium reduction performance (additional details are provided in the next section).
' Cycle time refers to the changing of the membrane operating mode from filtration to resting. In nitration mode, a vacuum is maintained to pull water through the membrane and produce treated effluent. In resting mode, no vacuum is applied to the membranes and the air scour is maintained to help remove solids. ( I ]
4-5
Results and Discussion
Treated Water Characteristics This section discusses the characteristics of the MBR pilot treated effluent (permeate). The MBR process was operated to remove ammonia, nitrates, and selenium while being monitored for potential impacts from chemical scaling and high TDS. Membrane performance and optimization of the HRT and SRT were monitored throughout the study. Table 4-3 summarizes the treated effluent water quality based on samples collected during each operating condition. Figures 4-1,4-2, 4-3, 4-4, and 4-5 present MBR influent and effluent COD and phosphorus; ammonia and total Kjeldahl nitrogen (TKN); nitrate/nitrite nitrogen; selenium; and mercury, respectively. As seen in Table 4-3, the MBR effluent phosphorus concentration was higher during Condition I; however, as Figure 4-1 shows, the influent phosphorus concentration was consistent for the pilot study. The increased MBR effluent phosphorus concentration during Condition I was due to the higher-than-planned-for phosphorus content of the nutrient balanced acetic acid blend being added. Once the acetic acid was switched to the 80 percent solution in the first week of December 201 5 (near the beginning of Condition 2), the MBR effluent phosphorus concentration decreased.
4-6
Results amI Discussion
Table 4-3 MBR Treated Effluent (Permeate) Water Quality1
Parameter Units
Condition 1
Min
Max
Avg.
T o ta l S elenium
D issolved S elenium
ng/L
120
350
200
rtg/L
120
340
190
N O 2 /N O 3 -N Total M ercury
m g/L ng/L
0.19 1.44
160 3.94
35 2.46
D issolved M ercury
ng/L
1.10
3.33
2.02
Total A rsenic
D issolved A rsenic Total C alcium
D issolved C alcium
T o ta l M agnesium D issolved M agnesium T otal S odium A m m onia C onductivity
COD
ng/L ng/L m g/L m g/L
m g/L
m g/L
m g/L m g/L fiS /cm m g/L
3.7 3.6
440 450
1,900
2 ,0 0 0
0.31 13,000 260
5.1 5.0
520 510
2,500
2,500
13.0 16,000 640
4.7 4,5
476 476
2 ,1 0 7
2 ,1 3 8
3.75 13,923 501
Condition 2
Min
Max
Avg.
110
260
173
Condition 3
Min
Max
Avg.
170
590
337
Condition 4
Min
Max Avg.
170
460
231
120
260
167
170
600
330
160
460
233
0.12 1.60 0.23
0.24 3.68 3.06
0.15 2.58 1.96
0.13 1.01 0.64
21 5.67 2.97
4.0 2.67 1.92
0.09 1.93 0.66
0.18 1.94 0.90
0.13 1.94 0.78
4
5
4
-
-
-
-
-
-
4
5
4
-
-
-
-
-
-
420
530
456
440
490
454
460
470
465
420
540
460
430
480
457
460
470
465
1,700
2 ,1 0 0
1,840
-
-
-
-
-
-
1,700
2 ,1 0 0
1,840
-
-
-
-
-
-
420 4
11,000 370
420 14 14,000 570
420 9
12,429 513
410 0.23 11,000 370
500 12.0 13,000 550
476 5.09 12,500 433
500 1.3 12,000 370
500 110 13,000 570
500 37 1 2 ,2 0 0 455
4-7
Results and Discussion
Table 4-3 (continued) MBR Treated Effluent (Permeate) Water Quality1
Parameter Units
Condition 1
Min
Max
Avg.
Condition 2
Min
Max
Avg.
Condition 3
Min
Max
Avg.
BOD
m g/L
-
-
TOC
m g/L
-
-
F lu o rid e
m g/L
21
25
S ulfate
m g/L
-
-
O rtho
m g/L
13
36
P hosphate
-
<2.0
<2.0
<2.0
<2.0
<2.0
<2.0
--
14
14
14
6
69
37
23
22
27
24
22
25
24
-
6,700
6 ,9 0 0
6 ,8 0 0
6,000
9 ,1 0 0
7 ,1 0 0
19
1.2
1.2
1.2
-
-
-
T o ta l
m g/L
2
25
14
1.2
6.5
3.9
0.86
4.30
2.64
P h o sp h o ru s
TKN S ulfide TSS TDS Total A lkalinity B icarbonate A lkalinity
m g/L m g/L m g/L m g/L m g/L m g/L
32 <0.22
14,000
200 200
94 <0.22
20,000
200 200
67 <0.22
17,286
200 200
73 <0.22
4 12,000
100 100
100 <0.22
4 16,000
110 110
81 <0.22
4 14,714
105 105
59 <0.22
2 13,000
40 40
78 <0.22
6 16,000
180 180
69 <0.22
4 14,600
116 116
C arbonate
m g/L
-
-
A lkalinity
-
<3.9
<3.9
<3.9
<3.9
<3.9
<3.9
Notes: 1Average soluble concentrations were higher than total for some results due to analytical imprecision. gS/cm ^ microSiemen per centimeter Avg. = average Max = maximum Min - minimum --- no data
Condition 4
Min
Max Avg.
<2.0
<2.0
<2.0
2.5
3.8
3.0
25
27
26
6 ,3 0 0
7 ,9 0 0
6 ,9 1 7
-
-
-
0.03
3.00
2.06
53 <0.22
3 14,000
220 220
96 <0.22
6 16,000
270 270
71 <0,22
5 14,600
240 240
<3.9
<3.9
<3.9
4-8
Results and Discussion
Figure 4-1 also shows thal the MBR influent and effluent C<)D concentrations were similar, (iiven this, and the readily biodegradable nature o f acetic acid, the data suggest that the influent ( O l ) was non-biodegradahle. Also, from Table 4-4. it is evident that all the CO D added through acetic acid was consumed in the MBR since the M B R effluent BOD and TOC concentrations were very low (higher BOD and TOC would indicate potential for additional biological activity in the oxidation o f organic matter and the amount o f carbon in organic material, respectively). I his also raises the question whether all the C O D in the influent reported is contributed by analytical interference or some other oxygen-demanding constituents, as opposed to organic carbon. It could also suggest that the organic compounds present were not fully oxidized in TOC tests (thus TOC readings would be lower than actual concentrations). The theorv on analytical interference is drawn from low TOC concentrations reported during period o f higher COD concentration measurements.
MBR COD and Phosphorus
- M)
25
2n Ec>
Ui
5 g
10 a
0
X
X
7' x .
X X
x.
'X X
X
'X
X -%% X
Date
MBR Inflimnt ('.0 0 OMBR Influent Phosphorus
* MBR Influent ('.Oil Addlion oMBP Effluent Phosphorus
Figure 4-1 MBR influent and effluent COD and phosphorus
MBR Effluent COD
The data in Figure 4-2 show complete nitrification during October and November 2 0 15. During this time, the influent ammonia concentration was high (ranging from 50 to I30 mg/L). Towards the end of C ondition l and into ( onditions 2 and 3. the influent ammonia concentration decreased to 10 to 20 mg/L, and nitrification appeared to stall (as seen by an increase in MBR effluent ammonia concentration).
4-0
Kesuli.v tnul Discussion
MBR Am monia 7 KN Species
\s
v
;x ;x
\
\
\
\
Dale
MBR m ueni Ammonia oMHR inuenl tk n aMHR Effluent Ammonia
Figure 4-2 MBR influent and effluent ammonia and TKN
aMBR f (fluent TKN
Also during litis time, it was noticed that the pilot feed water alkalinity was lower than at the beginning ol the demonstration and higher ammonia concentrations were coming from the P4 boiler units and creating higher pH values.I
I he generation o f MLSS in the MUR varied throughout the pilot study. Wasting solids (M LSS) from the M UR was initiated during the middle o f Condition 2 on December 28. 2015. at a rate of 80 (.PI) and was stopped in the middle o f Condition 3 on January 30. 2016. I his resulted in decreased MLSS concentrations during Conditions 2 and 3 from approximately 6.000 mg L down to approximately 3,000 mg L and an increase in MLSS concentration during Condition 4 to approximately 4.200 mg L. These changes were not the cause o f the loss o f nitrification as the loss o f complete nitrification occurred on November 30, 2015. one month prior to initiating the solids wasting. I Itese changes do not appear to have impacted denitrification as full reduction o f nitrate-nitrogen to nitrogen gas occurred throughout the pilot study. During ( ondilion 4. the influent ammonia concentration started to increase with a corresponding increase in MBR effluent ammonia concentration, further indicating that nitrification had been lost. The reason(s) for losing the nitrification in the MBR could not be identified, but suspected to be caused by toxic or inhibitory compounds in the water.
4-11)
Results and Discussion
In order to understand if toxicity/inhibitory compounds were contributing to the loss in nitrification, potassium sulfite was dosed to the Equalization Tank starting January 5, 2016, at a rate of 50 mg/L (to achieve 25 mg/L after dilution with service water). Higher dosages were added throughout Condition 3. The potassium sulfite was added to reduce potential oxidants that may have been contributing to the loss in nitrification performance. There were also concerns that the reduction in nitrification performance indicated the presence of inhibitory or toxic compounds that could have potentially also affected the ABMet bacteria and selenium reduction performance.
Nitrification performance did not noticeably improve when dosing potassium sulfite and dosing was discontinued on January 27. In an attempt to rule out the effectiveness of potassium sulfite versus the presence/impact of inhibitory compounds, 20 kg of powdered activated carbon (PAC) was added to MBR on February 6, 2016. The PAC should have absorbed inhibitory compounds present in the water; however, MBR nitrification performance did not improve with PAC addition.
Another observation from Figure 4-2 is that the blended influent contained recalcitrant TKN concentrations averaging 65 to 81 mg/L during all four conditions. TKN is comprised of ammonia and organic-nitrogen and the high concentrations in the MBR effluent suggest that a significant portion of the organic nitrogen fraction of the influent TKN was in a non-biodegradable form.
Nitrate/Nitrite
One of the main goals of the pretreatment MBR was to remove (denitrify ) the nitrate/nitritenitrogen present in the influent and produced via nitrification in the aerobic zone. The average effluent nitrate/nitrite-nitrogen concentration during Condition 1 was 35 mg/L, indicating incomplete denitrification. This was due to the acetic acid dosing point inadvertently being installed in the aerobic zone of the MBR, which resulted in poor denitrification caused by anoxic zone #3 being carbon limited. After three weeks of poor denitrification, this situation was discovered and the acetic acid dosing point was relocated to the planned location (i.e., anoxic zone #3) on November 1I, 2015.
As can be seen in Figure 4-3, once the acetic acid dosing point was relocated, complete denitrification occurred. Complete denitrification continued in the MBR throughout the remainder of the pilot study (other than a 2-week period towards the end of Condition 3, when the acetic acid dose was reduced intentionally to allow nitrate-nitrogen to pass through the pretreatment MBR to the ABMet system). Although nitrification was lost during Conditions 3 through 4, nitrate-nitrogen removal (denitrification) was never a problem in the MBR. For some unknown reason, more sensitive nitrifying bacteria were inhibited but the heterotrophic denitrifying bacteria were not inhibited.
Result'' and Discussion
Condition 1
MBR Nitnte'Nitrate as N
Condition 2
Condition 3
Condition 4
MBR Influent M02/N03
Figure 4-3 MBR influent and effluent nitrate/nitrite as nitrogen
BMHR FIBuent R 02'N03
Selenium
As seen in figure 4-4, significant selenium reduction occurred in the pretrealmenl MBR. I he overall average iutluent total selenium concentration was 1.1 75 pg/l.. and the overall average effluent total selenium concentration was 250 pg/l On average throughout all conditions, approximately 75 percent reduction in selenium concentration was observed. I his is attributed to maintaining a verv low ORP (approximate!) -350 m V) in the anoxic zone #3 with high acetic acid dosage. This ( >RP is comparable It' that maintained in the ABMel reactors. Another observation from f igure 4-4 is that almost all of the influent selenium was soluble, which indicates that removal across the pretreatment MBR was hv biological reduction to elemental selenium (which exists in particulate form) followed bv nitration via the ultrafiltration membrane.
4-12
M BR Selenium
Rexults inici D iscu ssion
MHR Influent Total Selenium MBR Effluent Total Selenium
OMHR Influent Dissolved Selenium OMHR Effluent Dissolved Selenium
F ig u re 4-4
MBR influent and effluent selenium
On January I I. 201 (>. the MBR flow rate was increased from I GPM to I .ft (iP M . Prior to making this change, the MBR effluent was sampled and the selenium result was higher than previous results(220 pg/L \s. 120 pg I front January 7). Acetic acid (nutrient) teed was increased proportionally with the feed rate and no changes in ORP were observed: there were no immediate indications as to why the effluent selenium concentrations began to increase.
Higher selenium concentration results continued until midway through Condition 3. when they decreased hut again increased until the start o f Condition 4. The increase in diluent selenium at the beginning o f Condition 3 is not likely to be caused solely by the increased flow rate since the change occurred after the initial sample was taken that prov ided the indication. Selenium concentration results from the ABMet I and 2 reactor effluents were also higher that day (with no changes observed in either ( >RP or pH): again, emphasizing the increased flow rale was not the sole contributor.
Mercury
As seen in figure 4-5, significant mercury reduction also occurred in the pretreatment MBR. The average influent total mercury concentration to the pretreatment MBR was approximately 72 ng/L. and the average effluent was 3 ng/L. The pretreatment MBR stage achieved approximately 96 percent reduction o f mercury throughout the pilot study I his is likely due to mercury entering the MBR in mostly particulate form (see I able 4-1). the particulate mercury being removed by the ultra filtration membrane, and potentially due to some o f the soluble mercury precipitating to a particulate with sulfide generated under low ORP in the anoxic zone #3.
4-13
Results am i Discussion
MBR Mercury
MBR Influent Total Mercury MBR EtTiuent Total Mercury
Figure 4-5 MBR influent and effluent mercury
OMBR Influent Dissolved Mercury MBR Efllut- t Dissolved Mercury
Arsenic
Several weeks' worth o f arsenic data were collected, which indicated no removal o f arsenic in the pretreatment MBR. Average influent and effluent concentrations were both approximate!} 4.6 pg/L. I his is indicative o f the arsenic entering the M BR in mostly dissolved form, and not precipitating to a particulate form in ihc MBR (see I'ahle 4-1.
ABMet
Operating Parameters
ABMet operational parameters are discussed in this section. I he MBR effluent (permeate) was collected in the ABMet feed tank. From the feed tank, the water was pumped to the M3Met Stage I . Table 4-4 compares the ABMet operational parameters at the four test conditions.
4-14
Results and Discussion
Table 4-4 ABMet Operational Parameters
Condition Description
Condition Number Parameter/Dates
Startup and Acclimation
1
10/26/15 to 12/16/15
Initial Operation
2
12/17/15 to 1/10/16
Increased Flow/ Reduced HRT
Process Adjustments
3
4
1/11/16 to 2/13/16
2/14/16 to 3/3/16
A B M et S tage 1 Flow A verage A B M et Stage 1 EBC T A B M et S tage 2 F low A verage A B M et Stage 2 EBC T
0.5 GPM 11.2 hours 0.5 G PM 11.2 hours
0.6 G PM 9.4 hours 0.6 G PM 9.4 hours
0.9 G PM 6.2 hours 0.9 G PM 6.2 hours
0.9 GPM 6.2 hours 0.9 GPM 6.2 hours
A B M et Influent ORP A ve ra g e
183 mV
91 m V
94 m V
368 mV
Range
110 to 263 m V
52 to 151 m V
34 to 224 m V
261 to 454 m V
A B M et E ffluent O R P 1 A ve ra g e
-125 m V
-184 m V
-365 m V
-350 m V
Range
-500 to 106 mV -364 to 295 mV
B ackw ash Flow Rate
80 GPM
80 GPM
D egas Flow R ates
100 GPM
100 GPM
Feed Tank HRT
4 hours
4 hours
C hem ical Doses
A ntisealant, M D C 776 (M BR P erm eate to A B M et Feed Tank)
none
5 m g/L
-419 to -310 m V 80 GPM 100 GPM 4 hours
5 m g/L
-412 to -318 mV 80 GPM 100 GPM 4 hours
5 m g/L
1ORP data varied significantly, including positive values, potentially due to analytical error with the inline probes. EBCT = empty bed contact time, calculated as total boundary volume occupied by the GAC media divided by the
flow rate passing through the vessel
The average flow rate to the ABMet system during Condition I was 0.5 GPM, which corresponds to a total empty bed contact time (EBCT) of approximately 22.4 hours (1 1.2 hours for each stage). During Condition 2, the average flow rate increased to 0.6 GPM (due to mechanical issues, not intentionally increased), which corresponds to a total EBCT of approximately 18.8 hours. During Conditions 3 and 4, an intentional increase in flow rate to 0.9 GPM occurred. This corresponds to an average total EBCT of 12.4 hours. The project originally intended to increase flow rates to achieve EBCT closer to GE's target design of approximately 4 hours per reactor (8 hours total).
4-15
Results and Discussion
1he average ORP in the ABM et effluent during each condition ranged from -125 to -365 ni V, with the lowest average values o f -365 and -350 m V occurring during Conditions 3 and 4. respectively. These correspond to increased acetic acid dosing during Conditions 3 and 4 and were within (il.'s target range o f -250 to -400mV. Fhe average ORP values during Conditions I and 2 may be artificially elevated due to several reported positive values that may have resulted from analytical error and/or issues with the inline measurement probes. A plot o f inline ORP data from the MBR permeate (ABM et feed), ABMet stage I effluent, and ABMet stage 2 effluent are presented in Figure 4-6. Spikes in ORP from the ABMet locations are primarily associated with backwash and degas events. The ABM et system was taken offline from December 24 through 28 due to high TM P in the MBR system which required a cleaning. Both stages returned to <0 mV approximately one day after startup.
ORPl
Jl'toxls
-.on 10
Figure 4-6 ABMet inline ORP trends Courtesy of GE Water & Process Technologies
-Jan-16
U-<b-lS
-Mar.lt
l our different nutrients were trialed for the ABM et pilot system. The initial nutrient was a blend o f acetic acid, ammonium, and phosphorus (the same blend used initially for the MBR pilot). Due to formation ofexcess solids on the top o f each ABMet reactor, a second nutrient containing 80% acetic acid (20% water) replaced the blended product. Three inches o f carbon-containing excess solids were removed from each reactor. The third w as a diluted blend o f the first blended
4-16
Results anil Discussion
product, made up of 50% pure acetic acid and 50% blended product. During the final phase of operation, molasses was added to the nutrient mix of pure acetic acid and blended product. A summary of the nutrient dosages and resulting COD concentration ranges are provided in Table 4-5.
Table 4-5 ABMet Nutrient Dosages
Condition Description Condition Number
Startup and Acclimation
1
Initial Operation
2
Increased Flow/ Reduced HRT
3
Process Adjustments
4
Parameter/Dates
10/26/15 to 12/16/15
12/17/15 to 1/10/16
1/11/16 to 2/13/16 2/14/16 to 3/3/16
N utrient
B lended acetic acid product containing acetic acid,
am m onium , and phosphorus
50/50 m ix o f pure acetic acid (80% )
and blended product
M olasses and 50/50 m ix o f pure
acetic acid and blended product
A cetic A cid (C O D ) - A B M e t S tage 1
A ve ra g e
28 m g/L
100 m g/L
76 m g/L
112 m g/L
Range
28 m g/L
87 to 175 m g/L
69 to 90 m g/L
90 to 134 m g/L
A cetic A cid (C O D ) - A B M et S tage 2
A ve ra g e
none
28 m g/L
37 m g/L
36 m g/L
Range
none
28 m g/L
28 to 70 m g/L
34 to 40 m g/L
The ABMet reactor head loss trends in Figure 4-7 provide indications of backwash and degas frequencies throughout the demonstration. Head loss is the water level above the media where a higher head loss is associated with an increased hydraulic resistance through the carbon media (and is indication of increased water level). For ABMet systems, increased head loss can indicate a need to backwash and/or degas the filters due to increased biomass and other solids or gases, respectively. Starting the week of January 11,2016, degassing frequency was based on a schedule of once every Tuesday and Thursday (after water samples were taken). The length of each degassing event was originally set to 30 seconds, but was changed in mid-February to 90 seconds. According to GE, 90 seconds is more representative of full-scale systems.
4-17
Results and Discussion
nthes
|
50
*5
60
BMaq* I
Headloss
a Sttqe?
5
SO
25
20
15
10
5
M- Nr-v *5
IS D ec-15
? JOT- I t
Figure 4-7 ABMet headloss trends Courtesy of GE Water & Process Technologies
? / JOT l i t
11 to o 16
?H r U
One backwash was completed for each ABMet stage during the demonstration. Stage I backwash occurred on January 25 and stage 2 occurred on January 2b. Analytical results from samples taken at the Backwash Waste l ank after two backwash events are prov ided in Table 4-6, Note the selenium values are reported in mg/I. as opposed to gg/l .
Table 4-6 ABMet Backwash Analysis
Sample Location Total selenium Dissolved selenium
and Date
(mg/L)
(mg/L)
NO2 /NO1 (mg-N/L)
Total Suspended Solids (mg/L)
A BM et Stage 1
78
1/25/2016
57
0.060
8 .0 1
A BM et Staqe 2
53
1/26/2016
51
<0.045
7 .5 '
'According to GE, solids in backwash were lower than expected, which may be due to sampling error.
Treated Water Characteristics
This section discusses the characteristics o f the ABMet pilot effluent. Table 4-7 summarizes the effluent water quality based on samples collected during each operating condition. Figures 4-8, 4-9, 4 - It), 4-1 1,4-12, and 4-13 present ABM et influent and effluent CO D and phosphorus; ammonia and TKN; nitrate/ntrite nitrogen; selenium; mercury; and arsenic, respectively.
4-18
Results and Discussion
Table 4-7 ABMet Stage 2 Effluent Water Quality1
Parameter
Units
Condition 1
Min
Max Avg.
Total S elenium
pg/L
7
D issolved S elenium
pg/L
5
110
36
76
27
S e(IV )2 S e(V I)2 MeSe2 SeCN2
N O 2/N O 3-N
pg/L pg/L pg/L pg/L m g/L
--<0.045
0.14
0,06
Total M ercury
D issolved M ercury
ng/L ng/L
0.50 <0.31
4.57 3.16
1.45 0.73
Total A rsenic
D issolved A rsenic
pg/L
1.1
4.0
2.0
pg/L
1.2
3.6
2.0
Total C alcium
m g/L
430
580
477
D issolved C alcium
m g/L
420
590
477
Total M agnesium
m g/L
1,900
2,400
2 ,1 5 0
D issolved M agnesium
m g/L
1,900
2,400
2 ,1 4 0
A m m onia
m g/L
11
86
49
Condition 2
Min
Max Avg.
3
6
5
3
7
5
Condition 3 Min Max Avg.
15
62
36
15
56
37
Condition 4
Min
Max Avg.
14
39
21
11
26
17
<0.045 0.41 0.23
0.13 2.25 0.59
0.081 1.21 0.42
0.29
51.8
<0.25
11.5
<0.25
1.3
0.52
6 9 .7 3
<0.045 0.060
0.37
1.20
0.45
1.20
13 5.7 0.9 18 0.036 0.84 0.76
0.8 <0.25 <0.25
2.1 <0.045
0.53 0.68
4.6 <0.25 <0.25 10.8 0.081 1.16 0.90
3.8 <0.25 <0.25
7.2 0 .0 4 6 0.85 0.79
<0.36
1,8
<0.36
2.0
0.9
<0.36
0.76
0.54
0.82
0.82
0.82
0.9
0.38
0.76
0.61
0.82
0.82
0.82
410
500
451
430
480
452
460
500
480
400
520
460
440
470
455
460
500
480
1,600 1,600
1,900 2 ,0 0 0
1,743 1,786
1,800 1,800
2 ,1 0 0 2 ,1 0 0
1,950 1,950
1,900 1,900
2 ,1 0 0 2,100
2,000 2 ,0 0 0
5
6
5
0.3
22
6.4
0.47
98
31
4 -19
Results and Discussion
Table 4-7 (continued) ABMet Stage 2 Effluent Water Quality1
Parameter
Units
Condition 1
Min
Max Avg.
Condition 2
Min
Max Avg.
Condition 3 Min Max Avg.
Condition 4 Min Max Avg.
C onductivity
pS /cm 14,000 16,000 14,500 11,000 14,000 12,250 11,000 13,000 12,500 12,000 14,000 12,400
COD
m g/L
290
1,600
674
150
540
383
260
370
335
340
600
427
BOD
m g/L
-
-
-
54
54
54
4.6
46
22.3
26
74
49
TOC
m g/L
-
-
-
19
19
19
2.0
12
8.3
0.73
0.93
0.86
C hloride
m g/L
-
-
-
2 ,7 0 0
2 ,7 0 0
2 ,7 0 0
2,600 3,300
2 ,9 4 0
2,700 2,900
2,800
F luoride
m g/L
16
24
21
23
24
24
23
25
24
26
28
26
S ulfate
m g/L
-
-
-
6 ,5 0 0
7,000
6 ,7 5 0
6,300 8,300
6 ,8 6 0
5,600 8,200
6 ,8 1 7
Total P hosphorus m g/L
0.59
3.5
2.2
2
2
2
1.3
5.4
4.1
0.75
7.90
5.36
TKN
m g/L
53
150
72
48
83
62
45
69
56
45
72
58
S ulfide TSS TDS
m g/L m g/L m g/L
<0.22 12
15,000
<0.22 44
18,000
<0.22 28
16,750
<0.22 11
4 ,9 0 0
<0.22 18
19,000
<0.22 13
12,725
<0.22
30
4.0
8.0
13,000 16,000
18 6.1 14,800
25 3.0 14,000
58 21 17,000
36 8.3 15,200
Total A lkalinity
m g/L
-
B icarbonate A lkalinity
m g/L
-
-
-
79
160
120
130
200
167
190
420
300
-
-
79
160
120
130
200
167
190
420
300
C arbonate A lkalinity
m g/L
-
-
-
<3.9
<3.9
<3.0
<3.9
<3.9
<3.9
<3.9
<3.9
<3.9
' Average soluble concentrations were higher than total for some results, due to analytical imprecision : Analyzed at Brooks Applied Labs: Se(IV) - selenite; Se(VI).selenate; MeSe - selenomethionine; SeCN - selenocyanate ' SeCN concentration max value during Condition 3 appears to be biased high when compared to total and dissolved selenium maximum concentrations
4-20
Jti'svlt.s hJ Discussion
Fiiuire 4-8 shows lhat there was little removal o f influent C( )I ) bv the A BM et as the ABMet influent and effluent C O D concentrations were similar: this trend was also observed in the prctroatmenl MBR system I he data again suggest that the influent (. O D was contributed by analytical interference or some other oxygen-demanding constituent, as opposed to organic carbon. The average ABMet influent and effluent C< >L) concentrations for all conditions together were approximately 482 and 4(i> mg/l., respectively, during the pilot test.
Condition 1
ABM et CO D and Phosphorus
Court*ion 2
Condition 3
Condition &
MUR F m .w n l C O D
k ABM et Stage 2 Effluent C O D
A B M e t Stage 2 Ltitum ii Pho sp horus
A B M o l Stage 1 E ffluent C O D
OMBFt f-tltuent Phosphom s
A B M et S ta g e 1 Cffl ienl * Added C O D & A B M et Stage t F ffluont Phosphoius
Figure 4-8 ABMet influent and effluent COD and phosphorus
A slight increase in phosphorus concentration in the ABM et can he seen at times in Figure 4-8 (presented on the secondary y-axis). This was due to the fact that the nutrient blended acetic acid (containing phosphoms) was used as a carbon source for the ABMet throughout the pilot study. Optimization o f effluent phosphorus concentration was not an objective o f this study. 1 he phosphorus concentration could be reduced by separating the carbon source from the phosphorus source.
Nitrate/Nitrite
As seen in Figure 4-9. T K N and ammonia were not significantly impacted in the ABMet. I he presence o f nitrate nitrite nitrogen in the ABM et influent only occurred during Condition I (when the acetic acid was being added to the wrong zone in the pretreatment M B R ) and for a few weeks during Condition 3 (when nitrate nitrite nitrogen was intentionally being maintained in the MBR effluent to provide the ABM et w ith electron donors) (Figure 4-10).
f t o tilts inti Discussion Condition 1
A B M et A rrm o n ia T K N Species
(Condition 2
Condition 3
Condilion \
MBH Effluent Ammonia ABMet Stage 2 Effluent Ammonia
OMBR Effluent ik n aABMor Stage 2 Effluent IK ti
F ig u re 4-9 ABMet influent and effluent ammonia and TKN
X-h'e
lABMet Stage 1 Effluent Ammonia
Condition 1
ABMet Nitrite/Mitrate as N
Condition 2
Condition 3
Condition 4
MBR Effluent H02UO3
AOLlet Staoe t Effluent N02N 03
F ig u re 4-10 ABMet influent and effluent nitrate/nitrite as nitrogen
ABMet Stage ? I fluent N02N03
4-22
Results and Discussion
Selenium Selenium removal in the AB M cl varied throughout the pilot study, as seen in f igure 1-11.
Condition 1
A B M et S elenium
Condition 2 Condition 3
Condition 4
MBR Effluent total Selenium ABMet Stage 1 Etlluenl Total Selenium ABMet Stage 2 Effluent Total Selenium
F ig u re 4-11 ABMet influent and effluent selenium
MBR Effluent Dissolved Selenium o ABMet Stage 1 Effluent Dissolved Selenium ABMet Stage 2 Effluent Dissolved Selenium
During Condition I . improved performance is observed as the system is acclimating and the bacteria arc being established. I he best performance occurred during Condition 2. when the influent average total selenium concentration to Stage I was 173 pg/L and the effluent average total selenium concentration from Stage 2 was 5 pg/L. This is the only condition where the target effluent total selenium goal o f f> pg/L (half o f the El G limit due to 1:1 dilution) was achieved.
Selenium reduction during Condition 3 declined as the average ABM et effluent total selenium concentration increased from 6 to 3<> pg/L.. I rom Table 4-7, it can be seen that the predominant selenium species in the ABM et effluent during Condition 3 were selenite and selenocyanate. Unfortunately, selenium speciation data were not available in the M B R effluent. A s described in the section on the MBR operating performance, the initial increase in effluent selenium from the ABMet coincided with flow rate increases (HR I and EBC'T decreases) in the MBR and ABMet systems. However, the samples w ith the resulting higher selenium concentrations were taken
prior to changing the flow rates that same day. On January 7, total selenium in the final ABMet
effluent was 5.0 pg/L while on January 11, the same day flow rates were increased, total selenium was 15 pg/L. Given the strong performance o f both systems leading up to this time, there were no indications that increasing the flow rate would have dramatic impacts on performance. Had the selenium result from January 11 been available, the project team may have decided to forgo increasing flow rates until the underlying issue(s) could be addressed.
4-23
Results and Discussion
Complicating the matter even further, results for total selenium from the ABMet effluent (stage two reactor effluent) from January 18, 19, and 21 failed We Energies' lab QA/QC checks, in these samples, the matrix spike recovery values were on the low end of the acceptable range, indicating the presence of analytical interferences. All matrix spike recoveries up to this point had been within the method control limits and the lab did not change anything in the sample preparation or analysis protocol. These were the first sets of results that failed matrix spike QC and We Energies' lab staff performed various tests in attempt to provide accurate analyses. On January 18, the We Energies lab staff who collected the samples noted a strong sulfide odor that had not been present in prior samples. The same odor was present in this sample for the duration of the test. This issue resulted in total selenium values being reported much lower than values later determined to be actual. Total selenium concentration results from these dates were originally reported as 13 pg/L, 11 pg/L, and 8.2 pg/L vs. 15 pg/L, 22 pg/L, and 31 pg/L reported after oxidizing the samples in the sample bottles using hydrogen peroxide. At the time, these lower values falsely indicated the ABMet had recovered from the unconfirmed upset and was tracking back to total selenium concentration of 6 pg/L or less. The project team decided not to alter the current pilot conditions until the accuracy of analytical results improved. By the time the analytical interferences had been overcome, the pilot system had been outside of the performance targets for three weeks. Comparisons of reported vs. verified selenium concentrations are provided in Table 4-8 and additional information on We Energies' evaluations is presented in Appendix B.
Table 4-8 Comparison of ABMet Effluent Selenium Concentrations During Period of Matrix Spike Recovery Failures
Sample Date
1/18/2016 1/19/2016 1/21/2016
Total selenium initially reported
(n g /L )
Dissolved selenium initially
reported
(n g /L )
13
12
11
10
8.2
7.7
Total selenium reported after improving matrix spike recovery
(n g /L )
15
22
31
Dissolved selenium reported after
improving matrix spike recovery
(n g /L )
Not measured
The sulfide odor that became present in samples taken after January 18 correspond to higher sulfide concentrations in the ABMet effluent (see Figure 4-12). The presence of sulfide is expected as bacteria naturally reduce sulfate to hydrogen sulfide gas. The decline in selenium reduction performance suggests that sulfate-reducing bacteria were present and may have become too predominant in the 2nd stage reactor.
4-24
ASM et 2nd Stage Reactor Effluent
Results and Discussion
ABMet Stage 2 Effluent Sulfide
ABMet Stage 2 Effluent Total Selenium
F ig u re 4-12 ABMet effluent sulfide and selenium
During Condition 3, several activities were done in attempt to improve performance in the ABMet. Reducing the ABMet flow rate was not implemented because the flow rate change was not believed to be a major contributing factor to the loss o f selenium reduction performance.
The following actions were taken to improve ABMet performance.
Potassium sulfite had been added to the Kqualization l ank in an attempt to target oxidants potentially impacting the MBR nitrifying bacteria. Reducing oxidants was expected to also improve ABMet bacteria performance.
20 kilograms o f PAC were added to the MBK in an attempt to adsorb any potential inhibitor compounds. Again, this was already part o f the MBR plan in case the potassium sulfite addition not improved nitrification in the MBR
Nitrate was allowed to pass through the MBR to the ABMet. by reducing the acetic acid feed to the MBR. to provide more electron acceptors.
4-25
Results am ! Discussion
While performance did improve slightly in Condition 4 (where the average ABM et effluent total selenium decreased from 36 pg L, to 21 pg/l ). the performance observed during Condition 2 and achievement o f the effluent total selenium goal o f 6 pg L were not attained during ( ondition 4 Ciiven the number o f variables that may have impacted performance in the ABMet that occurred relatively close together, it is difficult to definitively say why performance decreased, as it may have been any one of. o r a combination of, the variables and the trial concluded before complete recovery of performance occurred. Total and dissolved selenium concentration values from the ABMet stage I and stage 2 reactors are pro\ ided in f igure 4 13.
ABMet Effluent Selenium
Condition 1
Condition 2 Condition t
Condition 4
2`iO 4*---------------------------------------r4------------- 1------------------- -----------------
200
150
O
E
(I)
S9 % 8
50
* % 8 * 1 A
*1 ,A
> / -
\ *
'0,
w xw w sw s
Dale
ARMpt Stage 1 fc(fluent Total Selenium a ARWet Stage 2 Effluenl Total Selenium ---------- I argel t lf u e n i C .o n r e n t ia t o n (6 uq l )
O A H M rt Stag- t F flu e n t Dissok/ed Selenium AHWet Stag' 2 Effluent Dissolved Selenium
F ig u re 4-13
ABMet effluent selenium
As noted in discussion o f f igure 4-I (graph o f M B R influent and effluent C O D and phosphorus), influent and effluent CO D concentrations were similar, suggesting that the influent C O D was non-biodegradable. Jar tests with Kenton reagent were conducted on site to preliminarily investigate the feasibility o f Advanced Oxidation Processes (AOPs) to partially oxidize these non-biodegradable organics in wastewater [2]. However, the ja r tests were inconclusive due to high levels of inorganic scavengers, such as nitrate, sulfate, and chloride, and the limited time available. Simple onsite analytical tests were performed for CO D. h-N. N H 4-N , and pH on samples that had AOP (Fenton process) reactions as well as control samples, fhe results o f this testing did not show any clear trends, and were likely not very accurate as they were performed using Held test methods and were not performed in a controlled lab environment.
4 -2 6
Result.anti Discussion
Mercury
Further reduction o f mercury was observed in the ABMet data and can be seen in Figure 4 14. Mercury was reduced from an average o f 2.5 ng/L in the MBR effluent to I . I ng/L in the ABMet effluent. These averages are consistent with performance over the entire length o f the demonstration. The influent to the pilot system was compliant with the effluent mercury goal o f l7Xng/l (half o f the monthly I I Ci limit due to 1:1 dilution).
ABMet Mercury
w
0/5
MBR Effluent Total M ercury a ABMet Stage 2 Effluent Total Mercury
Figure 4-14 ABMet influent and effluent mercury
% Date
*e *0 ,6 *'6 *>6
o M B R Tffluent D issolved M ercury ABM et Stage 2 Effluent D issolved M ercury
4-27
Resull'i litui Discussion
Arsenic
As described abovc. arsenic was not removed in ihe MBR; however, as seen in Figure 4-15, removal o f arsenic in the ABMet did occur. The average ABM et influent arsenic concentration was 4.6 fig I . and the average effluent over the entire demonstration was 1.4 fig L, which is an average reduction o f approximately 70 percent. During Conditions 5 and 4. the average total arsenic in the effluent was 0.67 pg'l. (despite the loss o f performance in selenium reduction during this time). This is significant given the effluent arsenic goal o f 4 fig I due to the dilution with service water.
Condition 1
ABMet Arsenic
Condition 2
Condition 3
Condition 4
Pilot Influent Total Arsenic MBR Effluent Total Arsenic a ABM et Stage 2 Fftluent Total Arsenic
F ig u re 4-15
ABMet influent and effluent arsenic
J>8
o Pilot Influent D issolved A rsenic
a MBR Fftluent D issolved Arsenic
c AFiMef Stage 2 Fffluent D issolved Arsenic
4-2X
5
SUMMARY AND CONCLUSIONS
Pilot Sum m ary
A pilot demonstration of the GE MBR and ABMet processes ran from October 26, 2015 to March 3, 2016 (19 weeks). Results from the demonstration are described under four conditions that summarize performance changes observed. These conditions are summarized below.
Condition 1 - Startup and Acclimation (October 26, 2015 to December 16, 2015) - period where bacteria are grow ing and becoming established in the treatment system.
Condition 2 - Initial Operation (December 17, 2015 to January 10, 2016) - operational period defined by GE's original test plan.
Condition 3 - Increased Flow/Reduced HRT (January 11,2016 to February 13,2016)operational period intended to more closely simulate full-scale HRT/EBCT conditions.
Condition 4 - Process Adjustments (February 14, 2016 to March 3, 2016) - period where GE responded to decreased selenium removal performance of the ABMet.
Other than one complete system shutdown from December 24 through December 28 to perform a recovery cleaning on the MBR membranes, the system ran consistently through the duration of the demonstration.
The full-scale FGD wastewater treatment process at P4 has an average operating flow rate of 40 GPM and is typically operated five days per week to allow for equipment cleanings (scale removal from the secondary clarifier plates) on the weekends. The two P4 boilers burn Powder River Basin coal which allows their FGD systems to reach high cycles of concentration and therefore generates wastewater with high TDS and significant scaling potential. The FGD wastewater treatment process consists of a chemical precipitation system that includes hydrated lime addition; primary clarification; organosulfide, ferric chloride, and flocculation polymer addition; and secondary clarification. Two effluent discharge tanks after the secondary clarifier hold approximately 10,000 gallons each. The pilot system in this demonstration received water from the discharge tanks which was blended with plant service water (untreated water from Lake Michigan) in a small equalization tank prior to the MBR process.
GE Water & Process Technologies supplied and operated the ZeeWeed 500D MBR and ABMet
pilot systems. We Energies maintained operation of the physical/chemical treatment system.
After initial evaluations of the FGD wastewater, GE determined that blending with service water
(50/50 by volume) was required to minimize the impacts of scaling and TDS on the pilot systems
and bacteria performance. Removing scale-forming compounds (calcium, magnesium, and
sulfate) with chemical softening would have required an impractical amount of chemicals
(approximately 1,300 lbs./day) and produced high quantities of sludge (approximately
2,000 lbs./day).
"
"
"
"
5-1
Summary and Conclusions
The MBR process consisted of pre-anoxie, aerobic, and post-anoxic zones (to target denitrification [in the anoxic zones] and nitrification [in the aerobic zone]) and the ZeeWeed 500 membrane unit to filter biomass and other solids prior to the ABMet pilot. Permeate from the MBR was treated by two ABMet reactors in series. The MBR pilot was operated at I GPM (combined HRT of 35 hours) during Conditions 1 and 2 and at 1.6 GPM during Conditions 3 and 4 (combined HRT of 23 hours). The increase in flow rate was intentionally made by the project team to simulate full-scale conditions. Due to ABMet performance issues, the full-scale conditions for MBR HRT were not implemented. The MBR pilot ran consistently with only one shut down early in the demonstration for a membrane recovery cleaning. Following this cleaning, citric acid maintenance cleans were conducted every other day. Acetic acid products were used to supply nutrients (in the form of COD, ammonium, and phosphorus) to the MBR pilot. During Condition 1, a blended acetic acid product containing ammonium and phosphorus was used but resulted in excess solids scale formation. To mitigate this issue in Conditions 2 through 4, an 80% acetic acid (20% water) product was used in conjunction with a separate phosphoric acid product. The ABMet pilot was operated at 0.5 GPM (combined EBCT of 22.4 hours) during Condition 1, 0.6 GPM (combined EBCT of 18.8 hours) during Condition 2, and 0.9 GPM (combined EBCT of 12.4 hours) during Conditions 3 and 4. The increase in flow rate during Condition 2 was not intentional, but the increase in Conditions 3 and 4 were made by the project team in order to more closely simulate GE's commercial EBCT target of 4 to 4.5 hours per reactor (8 to 9 hours total). Due to ABMet performance issues, the full-scale conditions for ABMet EBCT were not implemented. Each stage of the ABMet pilot underwent only one backwash during the demonstration. Starting in Condition 3, degassing was conducted twice per week. As with the MBR process, acetic acid products were used to supply COD and nutrients to the ABMet reactors. For Conditions 1 and 2, the blended acetic acid product was used. A 50/50 mix of blended product and pure (80%) acetic acid product was used in Condition 3. Molasses was added to this mix in Condition 4. A summary of pilot influent, MBR effluent, and ABMet effluent water qualities for Conditions 2, 3, and 4 is provided in Table 5-1. The pilot evaluation failed to demonstrate that the ABMet system (with the feed water diluted by 50% and supplemented with the MBR ZeeWeed system) can meet the ELG limits on this water source. Limited information are available to explain the exact issue(s) related to the loss of selenium reduction performance and several variables were identified during the demonstration which may have impacted selenium removal performance.
5-2
Summary and Conclusions
Table 5-1 Summary of pilot results
Parameter
Units
Pilot Influent Min Max Avg.
MBR Effluent
Min
Max Avg.
ABMet Effluent Min Max Avg.
Total S elenium D issolved S elenium N O 2/N O 3-N Total M ercury D issolved M ercury Total A rsenic D issolved A rsenic
Total S elenium D issolved S elenium N O 2/N O 3-N Total M ercury D issolved M ercury Total A rsenic D issolved A rsenic
Total S elenium D issolved S elenium N O 2/N O 3-N Total M ercury D issolved M ercury Total A rse n ic D issolved A rsenic
f-ig/L ng/L m g/L ng/L ng/L ng/L ng/L
ng/L ng/L m g/L ng/L ng/L Mg/L Mg/L
Mg/L Mg/L m g/L ng/L ng/L Mg/L Mg/L
920 900 70 41 8.1 3.4 3.4
790 800 78 20
7 0.45 0.45
770 710 65 27 7.4 no data no data
1800 1900 160
90 18 8.1 8.1
1000 1000
88 71 27 4.9 4.7
1100 1100
95 76 12 no data no data
1071 1071
95 63 12.2 4.7 4.7
910 915 83 48 13 3.3 3.3
932 919 80 52 9.7 no data no data
Condition 2
110
260
120
260
0.12
0.24
1.60
3.68
0.23
3.06
4
5
4
5
Condition 3
170
590
170
600
0.13
21
1.01
5.67
0.64
2.97
no data no data
no data no data
Condition 4
170
460
160
460
0.09
0.18
1.93
1.94
0.66
0.90
no data no data
no data no data
173 167 0.15 2.58 1.96
4 4
337 330 4.0 2.67 1.92 no data no data
231 233 0.13 1.94 0.78 no data no data
3 3 <0.045 0.41 0.23 <0.36 <0.36
15 15 <0.045 0.37 0.45 <0.36 0.38
14 11 <0.045 0.53 0.68 0.82 0.82
6 7 0.13 2.25 0.59 1.8 2.0
62 56 0 .0 6 0 1.20 1.20 0.76 0.76
39 26 0.081 1.16 0.90 0.82 0.82
5 5 0.081 1.21 0.42 0.9 0.9
36 37 0 .0 3 6 0.84 0.76 0.54 0.61
21 17 0 .0 4 6 0.85 0.79 0.82 0.82
Pilot Objectives
ELG monthly avg. limit
Target based on 100% dilution
12
6
-
-
4.4
2.2
356
178
-
-
8
4
-
-
12
6
-
-
4.4
2.2
356
178
-
-
8
4
-
-
12
6
-
-
4.4
2.2
356
178
-
-
8
4
-
-
5-3
Summary and Conclusions
MBR Conclusions
The increased MBR effluent phosphorus concentration during Condition 1 was due to the higher-than-planned-for phosphorus content of the nutrient balanced acetic acid blend being added.
Influent and effluent BOD and TOC concentrations were low, indicating that all of the acetic acid added was degraded and this suggests that the influent and effluent COD concentrations of approximately 600 mg/L may be caused by analytical interference or some other oxygen demanding constituents as opposed to organic carbon.
Complete denitrification (reduction of nitrate nitrogen to nitrogen gas) was achievable in the pretreatment MBR.
Although nitrification was lost during Conditions 3 and 4, nitrate-nitrogen removal (denitrification) was never a problem in the MBR. For some unknown reason, more sensitive nitrifying bacteria were inhibited while the heterotrophic denitrifying bacteria were not inhibited.
Selenium, present mainly in dissolved form, was significantly reduced by the pretreatment MBR (approximately 75 percent on average for the entire demonstration).
Mercury, present mainly in particulate/insoluble form, was significantly reduced by the pretreatment MBR (approximately 96 percent on average for the entire demonstration).
Arsenic, present mainly in dissolved form, was not removed in the pretreatment MBR.
AB M et Conclusions
To EPRI's knowledge, this is the first pilot to attempt to treat FGD wastewater from a power plant burning 100% subbituminous coal.
The pilot test did not consistently demonstrate that the MBR/ABMet combination of technologies adequately treats selenium to meet the 2015 ELGs limits.
The flow rate to the ABMet was increased on January 11,2016 from an average of 0.6 GPM to an average of 0.9 GPM. This resulted in the combined EBCT in the ABMet decreasing from 18.8 to 12.4 hours. GE's target of approximately 8 hours for full-scale applications was a goal for the pilot demonstration but was never achieved.
The flow rate was also increased to the MBR on January 11,2016 from 1 to 1.5 GPM, which decreased the anoxic HRT from 20 to 13 hours. After this flow rate change (during Condition 3), the average MBR effluent total selenium increased by a factor of 2, doubling the selenium load on the downstream ABMet. However, analytical results taken prior to the flow rate increase were already showing an increase in ABMet effluent selenium. This information was not yet available to the project team when the decision was made to increase flow rates.
5-4
Summary and Conclusions
Another variable that may have impacted performance is the complete denitrification that was achieved in the pretreatment MBR. The low nitrate-N available to the ABMet may have resulted in the desired heterotrophic bacteria capable of reducing nitrate/nitrite and selenate/selenite being displaced by sulfate-reducing bacteria in the absence of nitrate-nitrogen as electron acceptor (ABMet average effluent sulfide increased from non-detect to 18 mg/L from Condition 2 to Condition 3 and increased sulfide odors were observed by onsite personnel). The high cycles of concentration associated with the P4 absorbers results in sulfate concentrations approximately five to seven times higher than FGD wastewater from units that burn eastern bituminous coals [3],
The presence of sulfate-reducing bacteria, indicated by the high concentration of hydrogen sulfide (H:S) in the second reactor effluent, may also have influenced the loss of selenium reduction. The high COD available to the ABMet from dosing acetic acid may have resulted in the desired bacteria, capable of reducing selenate/selenite, being outeompeted by sulfate-reducing bacteria in the presence of the highly reducing conditions.
Different nutrient product combinations were used in the demonstration (blended acetic acid containing ammonium and phosphorus, 80% acetic acid (20% water), an even mixture of the previous two products, an even mixture of the first two products and also molasses). Nutrient dosages were maintained based on COD concentrations, pH, and ORP and were within specifications throughout the demonstration. During Condition 4, GE began adding molasses in an attempt to mitigate excessive sulfate-reducing bacteria growth, but by the end of the demonstration (which ended due to a lack of FGD wastewater/pilot feed water supply) the target for selenium reduction had not been re-achieved. Acetic acid products may not provide benefits over molasses for selenium reduction (which is used at GE's commercial ABMet systems treating FGD wastewater for selenium), but there is insufficient information at this time to confirm this possibility.
Significant arsenic reduction (approximately 70 percent on average for the entire demonstration) occurred in the ABMet. The effluent goal of 4 pg/L was achieved. This is notable, as other EPR1 research has shown that for those few plants that have significant dissolved arsenic in their FGD wastewater, little removal is achieved by the physical/chemical treatment system [4J.
Summary and Conclusions
Recom m endations for Additional Research
As prev iously identified, many variables may have contributed to the loss of performance in the MBR nitrification process and the ABMet selenium reduction process. The following recommendations for additional research are made based on the results of this pilot demonstration.
Future studies should aim to identify compounds present in FGD wastewater that may be inhibitory or toxic to bacteria used for nitrification, denitrification, and metals reduction. EPRI suspects that oxidants naturally present in FGD wastewater are one potential source of inhibitory effects, but their identity and control have not been fully defined [5], Further studies to remove non-biodegradable organics in FGD wastewater are recommended. This demonstration tested the use of potassium sulfite, powder activated carbon, and advanced oxidation processes (Fenton reactions), but results are inconclusive.
The impacts of TDS on bacteria performance should be studied in a controlled environment. Testing should evaluate how increasing FGD wastewater TDS concentration could inhibit reduction of metals or slow down biochemical reactions. Variability in FGD wastewater chemistry is a consistent challenge and potential upsets to treatment technologies are a risk for regulatory compliance. For this demonstration, dilution of PCs chemical precipitation treatment effluent reduced the TDS from an average of 32,300 mg/L to 18,600 mg/L. The dilution was also required to mitigate the high scaling potential of this wastewater. Dilution may not be an option for every site.
The potential impacts of chemical scale formation in biological treatment FGD wastewater applications should be considered in future demonstrations. For this demonstration, dilution was used as a means to mitigate scale formation in the pilot systems, but dilution may not be an option for every site. General guidance for chemical softening are provided in the Wet Flue Gas Desulfurization Wastewater Physical/Chemical Treatment Guidelines [6],
The pilot MBR process demonstrated strong performance in reducing nitrate/nitrite prior to the ABMet, but with the potential for competing sulfate-reducing bacteria to negatively influence selenium reduction, it is suggested that nitrate/nitrite reduction as a pretreatment step be more closely evaluated. At this time, there is no clear guidance or rule of thumb for denitrification prior to selenium reduction for a specific range of influent nitrate/nitrite. Future studies that evaluate denitrification as a pretreatment step could consider bypassing a portion of the feed stream directly to the selenium reduction step because controlling denitrification to a target effluent concentration (such as 5 to 10 mg-N/L) may not be feasible or practical. Future studies should also evaluate if sulfate concentrations are also related to the loss of selenium reduction performance due to the influence of sulfate-reducing bacteria.
Three different nutrients were fed throughout the study. The use of the acetic acid products was a departure from GE's current commercial systems treating FGD wastewater and use molasses. The balance of COD, nitrogen, and phosphorus is important for maintaining bacteria performance in reducing metals. Future studies should focus on the control and application of nutrient products related to nitrification, denitrification, and selenium reduction performance for FGD wastewater.
Summary and Conclusions
The analytical interferences observed in the ABMet pilot effluent (stage two reactor effluent) should be further studied to determine ways to improve accuracy for total selenium analysis. It is hypothesized that the highly reducing conditions of the water resulted in selenium remaining in a reduced valence state (insoluble) and reporting higher than actually present. There have been instances from other tests (see Appendix B) where insoluble selenium was adsorbed to sample bottles and led to total selenium results reporting lower than actually present and con finned by dissolved species analyses.
FGD wastewater is impacted by upstream air emission control equipment, boiler operation, fuel selection, and many other factors. Changes in chemistry and flow rates are common for many sites and the wide range of conditions could impact performance of wastewater treatment systems. Coordination with FGD and upstream operations should be considered for any FGD wastewater treatment application, including biological treatment. However, there are limits to what can be achieved by coordination with upstream systems. The final treatment system should be robust enough to handle system variability in a balanced and coordinated approach. Additionally, physical/chemical treatment systems upstream of biological treatment can have an impact on the reliability of selenium reduction and are an integral part of the treatment process.
Limited data are publicly available on the operation and performance of biological treatment systems for FGD wastewater. The 2015 U.S. EPA Effluent Limitations Guidelines and Standards for the Steam Electric Power Generating Point Source Category may result in many new commercial systems over the next two to seven years. Continued research is needed to ensure power plant operators have the information they need to design, operate, and maintain treatment systems.
5-7
6
REFERENCES
1. Using Membrane Bioreactors for Wastewater Treatment in Small Communities. GE Power & Water, Trevose, PA. 2012. TP1037EN.
2. Dang, Yang, and Renzun Zhao. Advanced Oxidation Processes (AOPs) in Wastewater Treatment. Curr Pollution Rep, Sept. 2015, p. 167-1 76.
3. Guidance Document for Treatment o f Flue Gas Desulfurization (FGD) Wastewater in
Compliance with the 2015 Effluent Limitations Guidelines (ELGs). EPRI, Palo Alto, CA:
2016.3002008318.
"
4. Flue Gas Desulfurization (FGD) Wastewater Chemical Precipitation Treatment Performance Characterization Study. EPRI, Palo Alto, CA: 2015. 3002006251
5. Pilot Evaluation o f the A B M ef Biological Treatment System for Selenium Reduction in Flue Gas Desulfurization Wastewater Treatment. EPRI, Palo Alto, CA: 2014. 3002004550.
6. Wet Flue Gas Desulfurization Wastewater Physical/Chemical Treatment Guidelines. EPRI,
Palo Alto, CA: 2016. 3002008515.
'
6-9
A
SAMPLING AND LABORATORY ANALYTICAL PLAN
A-l
Sampling and Laboratory Analytical Plan
Table A-1 Analytical Plan - Frequency of Samples
Parameter
A m m onia A rsenic, total
A rsenic, filtered
C alcium , total
C alcium , filte re d C onductivity, p S /c m COD F luoride, total M agnesium , total M agnesium , filte re d N itra te /N itrate (N ) P h o sp h o ru s, total S elenium , total
Laboratory
TestA m erlca We E n e rg ie s We E n e rg ie s We E n e rg ie s We E n e rg ie s TestA m erica
TestA m erica TestA m erica We E n e rg ie s We E n e rg ie s TestA m erica
TestA m erica
We E n e rg ie s
Method
SM4500 EPA 200.8 D R C -02 EPA 200.8 D R C -02 EPA 200.7
EPA 200.7
SM2510B
SM5220C SM4500 EPA 200.7
EPA 200.7
SM4500
SM4500
EPA 2 0 0 .7 */ 200.8 D RCNH3
Service Water (S1)
2/m onth 2/m onth 2/m onth 2/m onth 2/m onth 2/m onth 2/m onth 2/m onth 2/m onth 2/m onth 2/m onth 2/m onth 2/m onth
Phys/Chem Effluent (S2)
Inf/EQ Tank Effluent (S3)
Condition 1
2/m onth
3/w eek
2/m onth
1/w eek
2/m onth
1/w eek
2/m onth
3/w eek
2/m onth
3/w eek
2/m onth
3/w eek
2/m onth 2/m onth 2/m onth
3/w eek 3/w eek 3/w eek
2/m onth
3/w eek
2/m onth
3/w eek
2/m onth
3/w eek
2/m onth
3/w eek
MBR Mix Liquor (S4)
MBR Effluent
(S5)
ABMet 1st Stage Effluent
(S6)
ABMet 2nd Stage Effluent
(S7)
3/w eek 1/w eek 1/w eek 3/w eek 3/w eek 3/w eek 3/w eek 3/w eek 3/w eek 3/w eek 3/w eek 3/w eek 1/w eek
3/w eek 3/w eek 3/w eek 3/w eek 3/w eek 3/w eek
1/w e e k
1/w eek 1/w e e k 1/w eek 1/w eek 1/w eek 3/w eek 1/w e e k 3/w eek 3/w eek 3/w eek
1/w e e k
ABMet Backwash
(S8)
A-2
Sampling and Laboratory Analytical Plan
Table A-1 (continued) Analytical Plan - Frequency of Samples
Parameter Laboratory
Method
Service Water (S1)
S elenium , filte re d
S ulfite
S ulfide TDS TKN TSS
We E nergies
We E n e rg ie s
TestA m erica TestA m erica TestA m erica TestA m erica
EPA 200.7* / 200.8 D RCNH3
O n-site; a n a ly z e im m ediately
SM4500
SM2540C
SM4500
SM2540D
2/m onth 2/m onth
A lkalinity A lum inum , total
A lum inum , filte re d A m m onia A rsenic, total
A rsenic, filtered
B icarbonate BOD Boron, total
TestA m erica We E n e rg ie s We E n e rg ie s
TestA m erica We E n e rg ie s
We E n e rg ie s TestA m erica
TestA m erica We E n e rg ie s
SM2320B EPA 200.7
EPA 200.7
SM4500 EPA 200.8 D R C -02 EPA 200.8 D R C -02 SM2320B SM5210B EPA 200.7
1/m o n th
1/m o n th 1/m o n th 1/m o n th 1/m o n th
Phys/Chem Effluent (S2)
Inf/EQ Tank Effluent (S3)
Condition 1 (continued
2/m onth
3/w eek
MBR Mix Liquor (S4)
2/m onth
3/w eek
2/m onth
3/w eek
2/m onth
3/w eek
2/m onth
3/w eek
2/m onth
3/w eek
3/w eek
Conditions 2, 3 and 4
1/m o n th
2/w eek
1/m o n th
1/w eek
1/m o n th
1/w eek
1/m o n th 1/m o n th
2/w eek 1/w e e k
1/m o n th
1/w eek
1/m o n th 1/m o n th 1/m o n th
2/w eek 1/w e e k 1/w eek
MBR Effluent
(S5)
1/w eek 3/w eek 3/w eek 3/w eek 3/w eek
2/w eek
2/w eek
2/w eek 1/w eek
ABMet 1st Stage Effluent
(S6)
ABMet 2nd Stage Effluent
<S7)
1/w eek
1/w eek
1/w eek
3/w eek 3/w eek
1/w eek 1/w e e k 3/w eek 1/w eek
2/w eek
2/w eek
2/w eek 1/w e e k
2/w eek 1/w e e k
1/w eek
2/w eek 1/w e e k
1/w eek
2/w eek 1/w eek 1/w e e k
ABMet Backwash
(S8)
Sampling and Laboratory Analytical Plan
Table A-1 (continued) Analytical Plan - Frequency of Samples
Parameter Laboratory
Method
Boron, filtered
C adm ium , total
C adm ium , filte re d C alcium , total
C alcium , filte re d C arbonate C hlorides, filte re d C obalt
COD C onductivity, p S /c m C opper, total
C opper, filtered
F luoride, total Iron, total
We
E nergies
We
E n e rg ie s
We
E n e rg ie s
We
E n e rg ie s
We
E nergies
TestA m erica
TestA m erica
We
E n e rg ie s
TestA m erica
TestA m erica
We
E n e rg ie s
We
E n e rg ie s TestA m erica
We
E nergies
EPA 200.7
EPA 2 0 0 .7 / EPA 200.8 EPA 2 0 0 .7 / EPA 200.8 EPA 200.7
EPA 200.7
SM2320B SM4500
EPA 200.7
SM5220C SM2510B
EPA 2 0 0 .7 / EPA 200.8 EPA 2 0 0 .7 / EPA 200.8 SM4500 EPA 200.7
Service Water
(S1)
Phys/Chem Effluent (S2)
Inf/EQ Tank Effluent (S3)
MBR Mix Liquor (S4)
Conditions 2, 3 and 4 (continued)
1/m o n th
1/w eek
MBR Effluent
(S5)
ABMet 1st Stage Effluent
(S6)
ABMet 2 nd Stage Effluent
(S7)
1/w e e k
1/m o n th
1/w e e k
1/w e e k
1/m o n th
1/w eek
1/w e e k
1/m o n th 1/m o n th
2/w eek
2/w eek
2/w eek
2/w eek
1/m onth 1/m onth
2/w eek
2/w eek
2/w eek
2/w eek
1/m onth
1/m o n th 1/m onth
1/w e e k 1/w eek
1/w eek
1/w e e k
1/w eek 1/w eek
1/m o n th
1/w eek
1/w eek
1/m o n th 1/m o n th
1/m o n th 1/m o n th
2/w eek 2/w eek
2/w eek 2/w eek
2/w eek 2/w eek
2/w eek 2/w eek
1/m o n th
1/w eek
1/w eek
1/m o n th
1/w eek
1/w eek
1/m o n th
1/m o n th 1/m o n th
2/w eek 1/w eek
2/w eek
2/w eek
2/w eek 1/w e e k
ABMet Backwash
(S8)
Sampling and Laboratory Analytical Plan
Table A-1 (continued) Analytical Plan - Frequency of Samples
Parameter Laboratory
Method
Iron, filtered
Lead, total
Lead, filtered
M agnesium , total M agnesium , filte re d M anganese, total M anganese, filte re d M ercury lowlevel, total M ercury lowlevel, 0.45 pm M olybdenum , total M olybdenum , filte re d N ickel, total
N ickel, filtered
We E n e rg ie s
We E n e rg ie s
We E n e rg ie s
We E n e rg ie s
We E n e rg ie s
We E n e rg ie s
We E n e rg ie s
We E n e rg ie s
We E n e rg ie s
We E n e rg ie s
We E n e rg ie s
We E n e rg ie s
We E n e rg ie s
EPA 200.7
EPA 2 0 0 .7 / EPA 200.8 EPA 2 0 0 .7 / EPA 200.8 EPA 200.7
EPA 200.7
EPA 200.7 EPA 200.7
EPA 1631E EPA 1631E EPA 200.7 EPA 200.7 EPA 200.7 EPA 200.7
Service Water (S1)
Phys/Chem Effluent (S2)
Inf/EQ Tank Effluent (S3)
MBR Mix Liquor (S4)
Conditions 2, 3 and 4 (continued)
1/m onth
1/w e e k
MBR Effluent
(S5)
ABMet 1st Stage Effluent
(S6)
ABMet 2nd Stage Effluent
(S7)
1/w e e k
1/m o n th
1/w e e k
1/w eek
1/m onth
1/w e e k
1/w eek
1/m onth 1/m onth
2/w eek
2/w eek
1/m o n th 1/m o n th
2/w eek
2/w eek
1/m o n th
1/w eek
1/w eek
1/m o n th
1/w eek
1/w eek
1/m o n th 1/m o n th
2/w eek
2/w eek
2/w eek
1/m o n th 1/m o n th
2/w eek
2/w eek
2/w eek
1/m o n th
1 /w e e k
1/w eek
1/m onth
1/w eek
1/w e e k
1/m onth
1/w e e k
1/w e e k
1/m o n th
1/w e e k
1/w e e k
ABMet Backwash
(S8)
A-5
Sampling and Laboratory Analytical Plan
Table A-1 (continued) Analytical Plan - Frequency of Samples
Parameter Laboratory
Method
N itrate/N itrate (N) P h o sp h o ru s, total S elenium , total
S elenium , filte re d S ilver
S odium , total
S odium , filtered
S ulfate, filtered S ulfite
S ulfide TDS T hallium
T in
TKN TOC
TestA m erica
TestA m erica
We E nergies We E n e rg ie s We E n e rg ie s We E n e rg ie s We E nergies TestA m erica We E n e rg ie s
TestA m erica TestA m erica We E n e rg ie s We E n e rg ie s TestA m erica TestA m erica
SM4500
SM4500
EPA 200.8 D RC-NH3 EPA 200.8 D RC-NH3 EPA 200.7
EPA 200.7
EPA 200.7
SM4500 O nsite; a n a ly z e im m ediately SM4500 SM2540C EPA 200.8
EPA 200.7
SM4500 SM5310C
A-6
Service Water (S1)
Phys/Chem Effluent (S2)
Inf/EQ Tank Effluent (S3)
MBR Mix Liquor (S4)
Conditions 2, 3 and 4 (continued)
1/m o n th 1/m onth
2/w eek
1/m o n th 1/m onth
1/w e e k
MBR Effluent
(S5)
2/w eek
1 /w e e k
ABMet 1st Stage Effluent
(S6)
ABMet 2nd Stage Effluent
(S7)
ABMet Backwash
(S8)
2/w eek 1/w eek
2/w eek 1/w eek
Each O ccurrence
1/m o n th 1/m onth 1/m o n th 1 /m o n th
1/m onth
2/w eek 2/w eek 1/w e e k
2/w eek 2/w eek
2/w eek 2/w eek
2/w eek 2/w eek 1/w e e k
Each O ccurrence
Each O ccurrence
1/m onth
1/w e e k
1/w e e k
1/w eek
1/m onth
1 /w e e k
1/w eek
1/m o n th
1/m onth 1/m onth
2/w eek 1/w e e k
2/w eek
2/w eek
2/w eek 1/w eek
1/m o n th
1/m onth 1/m onth 1/m onth
1/m onth
1/m onth
1/m onth 1/m onth
2/w eek 2/w eek 1/w eek
1/w e e k
1/w eek 1 /w e e k
2/w eek 2/w eek
1/w e e k 1/w eek
2/w eek 2/w eek
1/w e e k
2/w eek 2/w eek 1/w e e k
1/w e e k
1/w eek 1/w e e k
Table A-1 (continued) Analytical Plan - Frequency of Samples
Parameter Laboratory
Method
TSS
TestA m erica S M 2540D
Zinc, total
Zinc, filtered
A rsenic, total
A rsenic, filtered
A rsenic S pciation, filte re d M ercury low level, total M ercury lowlevel, 0.45 pm M ercury low level, 0.1 pm
We E n e rg ie s
We E n e rg ie s
B ro o k s Rand B ro o k s Rand B ro o k s Rand
B ro o k s Rand B ro o k s Rand
B ro o k s Rand
EPA 1638 DRC EPA 1638 DRC EPA 2 0 0 .8 /6 0 6 0 EPA 2 0 0 .8 /6 0 6 0 EPA 2 0 0 .8 /6 0 6 0
EPA 1631E DRC
EPA 1631E DRC
EPA 1631E DRC
Sampling and Laboratory Analytical Pian
Service Water (S1)
Phys/Chem Effluent (S2)
Inf/EQ Tank Effluent (S3)
MBR Mix Liquor (S4)
Conditions 2, 3 and 4 (continued)
1/m o n th
2/w eek
2/w eek
1/m o n th
1/w e e k
MBR Effluent
(S5)
2/w eek
ABMet 1st Stage Effluent
(S6)
ABMet 2nd Stage Effluent
(S7)
ABMet Backwash
(S8)
2/w eek
2/w eek 1/w eek
Each O ccurrence
1/m o n th
1/w e e k
1/w eek
1/m o n th
1/m onth
1/m o n th
1/m onth
1/m onth
1/m onth
1/m o n th 1/m onth 1/m onth
1/m onth 1/m onth 1/m onth
1/m onth 1/m onth 1/m onth
A -1
Sampling and Laboratory Analytical Plan
Table A-1 (continued) Analytical Plan - Frequency of Samples
Parameter Laboratory
Method
S elenium , total
B ro o k s Rand
S elenium , filte re d
B ro o k s Rand
S elenium S pciation, filte re d
B ro o k s Rand
Notes: InfVEQ influent equalization Phys/Chem physical/chem ical
EPA 2 0 0 .8 /6 0 6 0
EPA 2 0 0 .8 /6 0 6 0
EPA 2 0 0 .8 /6 0 6 0
Service Water (S1)
Phys/Chem Effluent (S2)
Inf/EQ Tank Effluent (S3)
MBR Mix Liquor (S4)
Conditions 2, 3 and 4 (continued)
1/m onth
MBR Effluent
(S5)
ABMet 1st Stage Effluent
(S6)
ABMet 2nd Stage Effluent
(S7)
1/m onth
1/m onth
1/m onth
ABMet Backwash
(S8)
1/m onth
1/m onth
1/m onth
1/m onth
3/m onth
1/w eek
A-8
B
ANALTYICAL QA/QC DISCUSSION
Results for total selenium from the ABMet effluent (stage two reactor effluent) from January 18, 19, and 21 failed We Energies' lab QA/QC checks. In these samples, the matrix spike recovery values were on the low end of the acceptable range. Low matrix spike recovery could indicate selenium is adhering to the container walls or precipitating out of solution. All matrix spike recoveries up to this point had been within the method control limits and the lab did not change anything in the sample preparation or analysis protocol. These were the first sets of results that failed matrix spike QC and We Energies' lab staff performed various tests in attempt to provide accurate analyses. On January 18, the We Energies lab staff who collected the samples noted a strong sulfide odor that had not been present in prior samples. The same odor was present in this sample for the duration of the test. Samples collected at the site for dissolved and total arsenic and selenium at We Energies' lab were not preserved in the field. The lab filtered a portion of the non-preserved sample for dissolved metals analysis and then acidified the filtered sample with nitric acid to achieve pH less than 2 s.u. The acidified samples were well shaken prior to pouring off the aliquot for totals metals digestion6. It was hypothesized that the low ORP in the ABMet stage two reactor effluent contributed to selenium remaining in a reduced valence state (insoluble). To test this hypothesis, the lab added 30% hydrogen peroxide and re-analyzed the samples. Select data comparing various sample preparation methods are presented in Table B -l. Analyses following peroxide addition resulted in matrix spike recovery improvements and total selenium concentrations reported higher than initial analyses. It is noted that EPA Method 200.7 does not include the addition of peroxide or other oxidants to improve accuracy .
To accomplish the quick turnaround time requests, the lab kept digestion time to two to three hours in the hotblock.
B-l
Analtyicul QA/QC Discussion
Table B-1 ABMet Second Stage Reactor Effluent Selenium Analysis Using Hydroxide Preparation Method
Date
1/18/2016
Tot. Se (M9/L)
Dis. Se
(og/L)
13
12
Dilution
10
MS/MSD mean % recovery
16
Spike Level
(pg/L)
40
Analysis Comments
EPA 200.2 digestion: 1mL H N O 3 + 0.25 mL HCI
15
15
10
27
no
10
data
15
no
10
data
1/19/2016
11
10
10
-
-
10
14
40
re-analysis for confirm ation
116
5
10x dilution d irect from sam ple bottle +
5 pg/L bench spike
N /A
N /A
R e-digested w ith 3m L/50m L 30% H 2 O 2
6
40
EPA 200.2 digestion: 1mL H N O 3 + 0.25 mL
HCI
104
N /A
5 pg/L post-digest (bench) spike
9.4
no
50
data
10
no
10
data
N /A
N /A
1:5 serial dilution o f 10x d igestate
7
N /A
alternate digestion - capped
26
no
10
data
1 /2 1 /2 0 1 6
8.2
7.7
10
19
no
10
data
26
no
25
data
105
5
10x dilution direct from sam ple bottle +
5 pg/L bench spike
11
40
EPA 200.2 digestion: 1mL HNCb + 0.25 mL
HCI
32
40
digested at 10x dilution
32
40
digested at 25x dilution
31
no
10
data
120
80
R e-digested w ith 3m L/50m L 30% H 2 O 2
1 /2 5 /2 0 1 6
42
-
10
11
9.0
10
106
40
M odified EPA 200.2 digestion: 2m L 30%
H2 O 2 + 1m L HNOs + 0.25 m L HCI
19
40
n o n -H 2 C>2 tre a te d
MS = matrix spike MSD matrix spike duplicate
Samples taken during Condition 3 of the demonstration were sent to Brooks Applied Labs in Bothell, WA for speciation analysis. The lab used EPA 200.8 and a closed-vessel, polytetrafluoroethylene (e.g., TeflonTM) bomb oven digestion. Sample analysis was performed by triple quadrupole mass spectrometry (1CP-QQQ-MS) to report total and dissolved selenium. Speciation analyses were conducted for selenite (Se(IV)), selenate (Se(VI)), selenomethionine (MeSe (IV)), selenocyanate (SeCN), and selenosulfate (SeSCh). These data are presented in
B-2
Analtyicul QA/OC Discussion
Table B-2. With a high level of analytical accuracy, the speciated selenium products should add lip to the dissolved selenium value for each sample. While minor variation may be expected, the January 28 sample had significant variance - total reported as 17 pg/L, dissolved reported as 9.28 pg/L, the sum of reported dissolved species is 135 pg/L. The current theory for this condition is that selenium adsorbs to the sample bottle, causing the analysis for total selenium to report lower than dissolved. The lab attempted to digest the sample in the sample bottle using 10% nitric acid, but the results were the same. The selenium may have already adsorbed to the bottle and the effect was irreversible. LPRI has begun additional studies to evaluate methods to improve analytical results from FGD wastewater biological treatment effluent. These studies are considering sample container use, preparation techniques, and analytical methods.
B-
co
Anal/yical OA OC Discussion
Table B-2 ABMet Effluent Selenium Spciation Results
B ro o k s A p p lie d Lab
D ate
T ot. Se
1/12/16 1/14/16 1/19/16 1 /2 1 /1 6 1/26/16
16.6 32.6 25.5 30.2 30.4
D is. Se
16.3 35.9 25.5 15.9 27.6
D is. Se(IV )
1.79 B 4.41 7.61 5.89 10.8
D is. S e (V I)
1.56 B 11.5 4.77
0.250 U 0 .9 6 5
D is. MeSe
(IV )
0.250 U
0.385 B 0.250 U 0.250 U 0.497 B
1/28/16 2/2/16 2 /4 /1 6 2/9/16 2 /1 1 /1 6 2 /1 6 /1 6 2 /1 8 /1 6 2 /2 3 /1 6 2 /2 5 /1 6
17.0
9.28
51.8
23.3
12.5
8.16
22.6
9.69
5.94
12.6
13.0
9.47
8.48 B 6.28 U 1.18 B
14.2
9.26 B
4.55
18.0 4 .88 B
3.01
8.80
5.23
2.06 B
10.3
3.79 B 0.828 B
0.965 U 0.965 U 0.965 U 0.250 U 0.250 U 0.250 U 0.250 U 0.250 U 0.250 U
1.32 0.588 B 0.638 B
0 .3 9 8 0.154 B 0.250 U 0.250 U 0.250 U 0.250 U
Note: all data reported as pg/'L U = not detected B = analyte is found in the associated blank, as well as in the sample
D is. SeCN
7.27 no data
7.46 14.7 11.9 69.7 20.1 17.2 11.4 5.24 10.8 8.01 2.10 B 2.65
D is. SeS3
1.10 B 2.21 B
2.53 2.53 2.50 U
3.82 2.50 U 2.50 U 2.50 U 0.275 B
2.73 2.53 U 2.04 B 0.222 U
D is. U nknow n
Se
0.00 u
0.465 3.1 4.60 1.52 8.71 1.12 1.11 1.02 1.63 1.69 1.46
0.551 0 .8 1 4
Sum of D is.
S p e cie s 11.7 19 25.5 25.2 25.7 135.4 30 24.9 22.3 8.5 19.8 12.5 6.8 4 .3
W e E nergies Lab
Tot. Se
D is. Se
17
20
45
47
22
no data
31
no data
46
42
47
46
62
54
50
46
35
31
23
19
29
25
27
22
18
13
16
13
B-4
c
RAW DATA
Raw data from the pilot study are provided in the attachment.
C-l
Export Control Restrictions Access to and use of EPRI Intellectual Property is granted with the spe cific understanding and requirement that responsibility for ensuring full compliance with all applicable U.S. and foreign export laws and regu lations is being undertaken by you and your company. This includes an obligation to ensure that any individual receiving access hereunder who is not a U.S. citizen or permanent U.S. resident is permitted access under applicable U.S. and foreign export laws and regulations. In the event you are uncertain whether you or your company may lawfully obtain access to this EPRI Intellectual Property, you acknowledge that it is your obligation to consult with your company's legal counsel to deter mine whether this access is lawful. Although EPRI may make available on a case-by-case basis an informal assessment of the applicable U.S. export classification for specific EPRI Intellectual Property, you and your company acknowledge that this assessment is solely for informational purposes and not for reliance purposes. You and your company ac knowledge that it is still the obligation of you and your company to make your own assessment of the applicable U.S. export classification and ensure compliance accordingly. You and your company understand and acknowledge your obligations to make a prompt report to EPRI and the appropriate authorities regarding any access to or use of EPRI Intellec tual Property hereunder that may be in violation of applicable U.S or foreign export laws or regulations.
The Electric Power Research Institute, Inc. (EPRI, www.epri.com) conducts research and development relating to the generation, delivery and use of electricity for the benefit of the public. An independent, nonprofit organization, EPRI brings together its scientists and engineers as well as experts from academia and industry to help address challenges in electricity, including reliability, efficiency, affordability, health, safety and the environment. EPRI also provides technology, policy and economic analyses to drive long-range research and development planning, and supports research in emerging technologies. EPRI members represent 90% of the electric utility revenue in the United States with international participation in 35 countries. EPRI's principal offices and laboratories are located in Palo Alto, Calif., Charlotte, N.C.; Knoxville, Tenn.; and Lenox, Mass.
Together...Shaping the Future of Electricity
P ro g ra m :
Water Management Technology
2 0 1 7 Elecfnc Power Research institute [ERRI), Inc. All rights reserved. Electric Power Research Institute, EPRI, and TOGETHER...SHAPING THE FUTURE OF ELECTRICITY are registered service marks of the Electric Power Research Institute, Inc.
3002006089
Electric Pow er Research Institute
3420 Hillview Avenue, Palo Alto, C alifornia 94304-1 338 PO Box 10412, Palo Alto, C alifornia 94303-0813 USA 800.3 13.3774 650.855.2 121 * aslcepri@epri.com www.epri.com
r US POSTAG
ZIP 23219 02 4A 0000332771
.
HUNTONSr
!
W ILLIAMS
J
HUNTON & WILLIAMS LLP RIVERFRONT PLAZA EAST TOWER
0
M l EAST BVRD STREET RICHMOND VIRGINIA 23219-4074
13
The Honorable E. Scott Pruitt
Administrator
Oc
United States Environmental Protec ion Agency
D(oQ/)j
William Jefferson Clinton Building 1200 Pennsylvania Avenue, NW
Maile Code: 1101A
03
E
Washington, DC 20460
s
v
j
l
R oute
EPA I
U N IT E D STATES
POSTAL SERVICE r>
J o ; Pruitt,Scott
USPS TRACKING #
9114999944238451101695
Mailstop ARN
9114 9999 4423 8451 1016 95
Department: 1101A S igcon
49999442384511016!
-