LRWRP_Design_Report_Chapter_6.pdf

CHAPTER 6

EFFLUENT FILTRATION (TM 4)

6.1 Introduction

The Lompoc Regional Wastewater Reclamation Plant (LRWRP) currently discharges treated effluent to San Miguelito Creek about 400 yards upstream of its confluence with the Santa Ynez River.

During dry periods, the treated effluent is the only flow in either stream. Filtration is not currently provided as a component of the existing treatment plant, so the proposed new filters will be entirely new facilities. As discussed below, filters are required to meet new discharge standards. Brown and Caldwell held a workshop with LRWRP staff on September 16,2004, to discuss the concepts and criteria to be used for liquid phase treatment processes, including effluent filtration. This technical memorandum (TM) is based on concepts discussed at the September lbth meeting and subsequent evaluations.

I<ennedy/Jenks prepared a Master Plan for LRWRP that included the following recommendations relative to effluent filtration:

Install continuously-backwashing upflow sand filters; 32 modules in 8 cells.

Design for 9.1 million gallons per day (mgd) peak dry weather flow (PDWF) plus redundancy.

Provide coagulant chemical addition.

The estimated construction cost of the recommended effluent atration facilities was $4,750,000.

Brown and Caldwell reviewed and evaluated these recommendations during preparation of our response to the request for proposals (RFP) for this project. This TM provides details to address needed filtration process additions.

Objectives

As stated in the scope of work and as discussed at the liquid process workshop, the following objectives are to be addressed in this TM:

Evaluate filtration recommendations and costs presented in the Master Plan.

Evaluate process requirements for filtration.

Evaluate potential filtration options and their performance characteristics.

Recommend the preferred filtration alternative.

Provide design criteria, mechanical layout, and site space requirements.

Estimate the cost of the recommended alternative.

P:\_I'rqects\ 12651 7 - LOhfl'OC\\XI",Project-Rep<)rt\Ii133979 Ptclirninar). Drsihm Report FinaLdoc February 2005

Effluent Filtration (TM 4) 6-2

Flow and Water Quality Projections

6.3.1 Flow Rates

The design flow rates were established at the liquid process workshop and are:

Average dry weather flow (ADWF) rate PDWF rate

5.5 mgd

9.5 mgd

Wet weather flows in excess of 9.5 mgd will be stored in EHB and returned for treatment when the flows subside below the PDWF rate. Furthermore, additional diurnal equalization of secondary effluent will be provided in the existing aeration basin to reduce the plant flow at the filters (and subsequent disinfection process) to 5.5 mgd. The basis of design for the filters is for a flow rate of

5.5 mgd.

6.3.2 Permit Requirements

LRWRP discharge to San Miguelito Creek and the Santa Ynez River is regulated under the provisions of National Pollutant Discharge Elimination System Permit No. CA0048127/Waste Discharge Requirements Order No. 01-87, dated May 18,2001.

6.3.3 Filtration Objectives

, Effluent filtration is not required to meet permit limitations on total suspended solids (TSS). The permit limits for TSS are 30 milligrams per liter (mg/L) on a 30-day average basis, and 45 mg/L on a 7-day average basis. Filtration is required to assist in meeting the effluent bacterial k t and to augment the removal of particulate solids that potentially contain metals.

The design of effluent filters will be based on California Title 22 requirements for the production of reclaimed water. While the current project is not intended to produce reclaimed water, it is prudent to follow Title 22 requirements in the design and specification of effluent filters so that water reclamation is an option available to the City of Lompoc in the future. In addition, following Title 22 requirements allows the system to be sized on the basis of validated design criteria. These criteria are known to meet stringent performance requirements based on test work performed at U.C. Davis following standardized procedures.

Title 22 requires that the turbidity of the filtered effluent not exceed a daily average of 2 nephelometric turbidity units (NTU), 5 NTU more than 5 percent of the time within a 24-hour period, and 10 NTU at any time. Coagulation can be waived if the filtered effluent does not exceed 2 NTU, the filter influent turbidity is continuously measured, and automated chemical addition is available when the influent turbidity exceeds 5 NTU. Coagulation facilities must be provided under Title 22 for automatic use whenever the turbidity of the filter influent exceeds 5 NTU. Coagulation, flocculation and clarification facilities wdl be constructed as part of the metals removal process imrnehately upstream of the filters. These fadties will meet the requirements for production of reclaimed water as delineated in Title 22 documents.

February 2005

Effluent Filtration (TM 4) 6-3

6.4 Effluent Filtration

This section discusses the evaluation of various effluent filtration process elements.

6.4.1 Evaluation of Filtration Technologies

The existing treatment plant does not provide effluent filtration, so all recommended filters will be new. In response to the RFP, Brown and CaldweU recommended the use of cloth media disc filters in place of the continuous-backwashing upflow filters recommended in the Master Plan. However, there are a number of filter options available for secondary effluent filtration and these are briefly described below, along with the basis of our recommendation.

The following types of filters are discussed:

Granular media filters Conventional deep bed filters Traveling bridge filters Upflow deep bed, continuous-backwash filters Deep bed mono-media filters Pulsed bed filters

Cloth media filters

Membrane processes (e.g., microfiltration, nanofiltration, reverse osmosis) also are considered to be filters, but are not discussed in this evaluation because membranes provide a hgh level of treatment that is not necessary on the project.

Granular Media Filters. Granular media filters have been in use for many years and have several design configurations. These filter types include:

Conventional Deep Bed Filters. Conventional deep bed filters have been used for many years, and are based on rapid sand filters used for water treatment (modified for wastewater effluent applications). This type of filter was the first filter approved in California for Title 22 applications, and is the standard for comparison. The filters are custom-engineered and are not vendor-supplied. They consist of compartmentalized media beds that are from 30 to 72 inches deep, and may contain mixed media of sand, anthracite and, occasionally, garnet. Anthracite on top will have an effective size of 1.2 to 1.7 millimeters (rnrn) and lower density than sand.

Quartz sand below the anthracite will have an effective size of 0.45 to 0.65 mm and garnet (if used) will have an effective size of 0.25 to 0.45 rnrn. The filters remove particulate matter throughout the bed and require intermittent backwash with water supplemented by air-scour. Backwash rates are between 12 and 20 gallons per minute per square foot (gpm/ft2), depending on the media and whether air scour is provided.

Conventional deep bed filters provide excellent filtrate quality and have performed well for secondary effluent filtration. Their &sadvantages include the complexity of the piping and controls that are required, the lack of a manufacturer to call on when

P:\-Projects\l26517 - I~OhlPOC\\~'P\Proicct-l<eporr\I~l3i977 Prelimtnar). D e s ~ q Report 1:inal.doc February 2005

Effluent Filtration (TM 4) 6-4 problems develop, and the large backwash flow rates that require separate equalization basins. Headloss at the end of filter run is as much as 8 feet. Ths type of filter is expensive to design and construct, and is not recommended for the LRWRP project.

Traveling Bridge Filters. Traveling bridge filters (also known as automatic backwashing or ABFTM filters) also have been available for many years and are based on earlier water treatment filter designs. The traveling bridge filter uses a shallow bed of either sand or sand and anthracite. The media depth is between 12 and 30 inches depending upon the manufacturer, with the deeper beds recommended for reclamation facilities. The filter cells are long, narrow channels with the media compartmentalized in small cells by plastic or fiberglass dividers. A traveling bridge is rail-mounted on the filter walls and moves along the bed for backwashing of the media. Backwash pumps, skimmers, and backwash hoods are mounted on the bridge. Power to the bridge-mounted equipment is provided by a festoon cable extended from one end of the filter.

During a backwash cycle, the bridge traverses the length of the filter. A backwash pump draws liquid from the filter effluent channel and conveys it to several hlter cells via a backwash shoe to remove accumulated solids. The backwash water is then transferred to a backwash channel for disposal. Backwash is continuous at a low rate of flow and does not require equalization.

Traveling bridge filters are limited to a loading rate of 2 gpm/ft2 by the California Department of Health Services (DHS) for reclamation applications (although the filters can be loaded up to 5 gpm/ft' for less critical applications). For this reason, the surface area used by a traveling bridge filter is approximately two and one half times that of most other filter types. Also, the travehg bridge filter is more of a surface filtering media that shows increased tendency to clog and overflow. For these reasons, the traveling bridge filter is not recommended for the LRWW project.

Manufacturers include Infilco-Degremont (IDI), Eirnco, and Aqua-Aerobic Systems, Inc.

Upflow Deep Bed Continuous-Backwash Filters. Upflow deep bed continuous-backwash fdters and related filter designs are based on the principal of recirculation of sand meha to a washing compartment, where solids are removed from the media and wasted and the sand is returned to the bed. These filters have performed well in secondary effluent service and produce an equalized backwash flow. Flow through the filter bed is in upward direction, and the bed is about 40 to 80 inches deep. The media have an effective size of 0.65 mn and are relatively uniform. A cone inside the filter is used to redistribute the washed sand, whch is lifted to the washing compartment by an air-lift pump. The backwash flow averages about 3 to 5 percent of peak throughput flow capacity, and a higher percent of average flow.

There are a number of manufacturers of sirmlar filters, each with design modifications to avoid process and patent concerns. These include Glters manufactured by Parkson Corporation (Dyna~and@), Waterhk (supersand@), Westech En . eering, (Techna~and@), Fluidyne Corporation (~luidsandq, Andritz (HydroSandTand hshbrook Corporation (strata-sand@). Each of these filters has

Effluent Filtration V M 4) 6-5 been approved by DHS for Title 22 applications in California. Loading rates of up to 5 gpm/ft2 are allowed with appropriate redundancy.

The upflow deep bed continuous-backwash filter is not recommended for the LRWRP project due to concerns over its ability to respond to a wide range of solids loading rates. This type of filter also is subject to operational problems if grease enters the confined sand bed area and b d d s up over h e .

Deep Bed Mono-Media Filters. This filter alternative uses a larger size, uniform media that are placed at greater depths to acheve better solids removal. The effective size media are between 1.8 and 2.5 mm and depths in the range of 72 to 120 inches have been used. Hydraulic loading rates of up to10 gprn/ft2 have been reported, but the rate is limited in Title 22 applications to 5 gpm/ft2. The filter is frequently used for combined filtration and denitrification, and is manufactured by Tetra Technologies, Inc., as the Tetram Filter. This filter is not recommended for this project, unless denitrification is required.

Pulsed Bed Filters. The pulsed bed filter is a proprietary unit marketed by U.S.

Filter/Zimpro as the Hydro-ClearTM filter, and therefore must be sole-sourced. The filter uses a relatively shallow media depth of 10 inches with an effective sand size of

0.45 mrn. The small media size provides excellent particle removal, and the bed is kept clean by intermittent aeration (or pulsing) to prevent surface fouling and short filter runs. The filter is divided into cells that are backwashed at a rate of 12 gpm/ft2 on a rotating basis. A proprietary grease removal system can be provided that helps keep the filter from clogging in high grease applications. The filter can be loaded to 5 gpm/ft2 with additional redundancy. Backwash volume is in the range of 2 to 3 percent of total throughput. The pulsed bed filter is not recommended for the LRWRP because of its proprietary design and the fact that grease removal is not anticipated to be a major concern.

Cloth Media Filters. The cloth media disc filter employs a series of cloth-covered discs suspended in vertical orientation in a tank containing effluent. Each &sc is made up of six pie-shaped sections mounted vertically to a central header pipe which conveys filtrate from the discs to the outlet pipe. The filter disc is static during operation, and rotates only during the backwash process. Up to 12 discs may be mounted on the center column to increase capacity.

During operation, water enters the filter tank through the inlet and passes by gravity through the filter cloth. Solids in the influent are removed and accumulate on the surface of the cloth media. Filtered water exits the tank through the interior of the Qsc and the central column. Backwashing starts based on either increased liquid level in the tank, or on a timer signal. Heavier settleable solids fall to the bottom of the filter tank and are occasionally removed by pumping to the plant headworks.

There are no moving parts during filtration.

During backwashing, the disc rotates such that a suction mechanism fits over a portion of the disc and removes accumulated solids from both sides by liquid suction. Discs are backwashed in groups of two to maintain filtration throughput capacity. The discs slowly rotate, allowing the entire surface to be cleaned, until the

P:?-l'rojects\126517 - LOhfl'OC\\Xl'\Projectt1?eport\11)33979 Preliminst). Design Report Final.doc February 2005

Effluent Filtration (TM 4) 6-6 process is complete. Backwash water is directed to the plant headworks for retreatment.

DHS allows the cloth media disc filter to be loaded up to a hydraulic application rate of 3.5 gpm/ft2 in reclamation operation. The effective application rates are lower to account for unit sizing and redundancy.

The filter cloth needs replacement about once every 3 years, which is a relatively simple operation. A prefit "sock" is provided that fits over each segment of the disc and installs rapidly. Each filter tank can be refit with new media in an %hour shift or less.

The cloth media disc filter is the recommended alternative for the LRWRP project.

This filter is manufactured by Aqua-Aerobic Systems, Inc. and I h g e r , Inc.

Appendix A lists current Aqua Aerobic Cloth Media Disc Filter installations upstream of W disinfection.

6.5 Recommended Effluent Filter Design Criteria

Recommended design criteria for the effluent filtration process are provided in Table 6.5-1.

Table 6.5-1 Effluent Filter Design Criteria

Number of tanks"

Size of tanks, feet

Effective filter surface area, ft2

Hydraulic loadmg, gpm/ft?. per dayb

Number of discs per tank

Area per filter disc, ft2

Backwash volume, gpm

Parameter

3 (2 service/l redundant)

19.75 long x 9.2 wide x 11.5 deep

1,940

2.96 @ 5.5 mgd flow rate

53.8

11 5

Value

Number of units assumes one unit out of service during average conditions.

Based on 3 percent of throughput of filter

The normal lower particle size for effective removal by granular media or cloth media fdtration is 10

pm. Particles smaller than 10 pm can be removed if they are coagulated upstream of the filter or strained through a build-up layer of solids on the filter surface. The particle size distribution in secondary effluent has been studied and found to be a function of the type of biological process used. (See the paper by Narayanan et. al. provided in Attachment B.) The referenced study indicates that shorter SRT biological treatment processes produce a lvgher percentage of particles less than 10 pn in size and a corresponding decrease in W transmittance downstream of filtration.

Conversely, longer SRT systems produce a greater number of larger particles and a higher downstream UV transmittance.

P:\-Projects\126517 -LOh~OC\\4P\Projrct~11~p~rt\1~133979 I'rcliminary Design Report Final.dor February 2005

Effluent Filtration (TM 4) 6-7

The proposed activated sludge system proposed for Lompoc falls in the longer SRT category and is expected to produce particles that are more amenable to effective filtration. Additionally, coagulation, flocculation and clarification will be provided by the metals removal process to huther reduce the TSS concentrations. These processes are expected to maximize filter runs and produce water that is readily inactivated by W disinfection.

6.6 Description of Filter Operation

The cloth media filter operates based on control s~gnals from an operator-adjustable microprocessor. Details are provided in this section.

6.6.1 Normal Filter Operation

Duiing normal filter operation, influent enters the filter tank and passes through the cloth media by gravity into the disc assembly (unfiltered water is in the tank itself). Solids accumulate on the cloth media as the filter run progresses. A mat of collected solids accumulates on the exterior of the cloth media that increases head loss through the filter and raises the depth of water in the tank. The accumulated mat of solids is called a "Sch.utzdecke" in filtration technology and increases the effectiveness of the media in removing very small particles.

Normal Operation Filtered water (filtrate) enters the internal portion of the disc and flows to the central support cylinder through which it flows to the outlet. The liquid level in the filter tank is controlled by the elevation of an outlet weir. Filtrate flows over the outlet weir to disinfection. Heavier solids in the influent stream settle to the tank bottom. There is no mechanical movement of hlter components during normal filter operation, so equipment wear is minimal.

The DHS maximum filter hydraulic loading rate for this type of filter is 3.5 gpm/ft2 of effective media area during design flow conditions with one unit out of service. The filters are designed to provide the 2.96 gpm/ft2 hydraulic loading with one ater tank out of service. This meets the Title 22 filter testing recommendations published by DHS and developed at U.C. Davis. The third fiter tank is redundant for use when one of the other tanks is out of service. It is permissible to operate all three tanks in parallel at lower loading rates.

P:\-Projects\l26517 - LOMPOC\~T~\Pmject-Rep0rt\IO33979 Preliminary Design Repoa F'md.doc February 2005

Effluent Filtration (TM 4) 6-8

6.6.2 Backwashing

The backwash cycle is initiated by level in the tank or elapsed time. Accumulated solids on the cloth-covered discs are removed by a reverse flow of backwash water through the center core cylinders and discs. Backwash water is collected in a "backwash shoe" assembly and directed to a collection sump for return to the plant headworks. The backwash flow is provided by the backwash pumps. The discs rotate slowly during backwashing to allow each segment to be cleaned Backwash-sequentially. Filtration is continuous through the Shoe remainder of the disc area, so that filter operation is not interrupted during backwashing. Only the small portion of each disc is removed from service during the process.

,- Rotation / of Disc

Backwashing

Backwash B FIOW to Pump

The discs rotate during the backwashing process so that the entire disc is backwashed in one minute.

Four discs per iilter tank are backwashed at a time so the entire filter tank can be backwashed in 3 minutes. About one percent of effective media area is removed from service for backwashing at any one time. Filtered water is returned through the media into the shoe as backwash water to remove the accumulated solids on the media. The shoe is fit tightly enough to prevent unfiltered water from entering the backwash system.

6.6.3 Settling

Because the filter is quiescent during operation (no moving parts), larger particles will settle by gravity in the hlter tanks and not accumulate on the cloth media. This will increase filter runs, but results in Inlet sludge accumulation at the bottom of the tank.

hism material is periodically drained to the backwash sump by gravity. A mechanized valve is operated automatically on the basis of a timer to accomplish this cleaning operation. The settling compartment is sloped to a hopper bottom, and a slotted pipe is used in the hopper to prevent "coning" of the solids during removal.

,- Automatic / "ah

Removal of Settled Solids Sump

6.6.4 Service

The filter cloth needs to be replaced on each disc about once every 3 to 5 years, depending on performance. The discs do not need to be removed fmm the tank, but the tank must be removed from service during this time. A hoist rail is provided over each tank for any major maintenance, although normal servicing does not require lifting. The central disc bearings may need replacement every 5 to 10 years, and this requires lifting of the rotating module. Other maintenance items include pump repair and typical routine servicing.

P:\_Projects\l26517 - LOhlPOC\\Y'P\Project-Report\I033979 Preliminary Design Report Finddoc February 2005

Effluent Filtration (TM 4) 6-9

6.7 Materials of Construction

The cloth media disc filter is available in several installation options, which include:

Equipment installed in concrete basins.

Equipment installed in painted carbon steel tanks.

Equipment installed in stamless steel tanks.

The budgetary equipment costs provided by Aqua-Aerobic Systems, Inc. for each option are shown in Table 6.7-1 on the following page.

Table 6.7-1. Budgetary Equipment Costs

Effluent filters; concrete basins

Effluent filters; stainless steel tanks

Effluent filters; painted steel tanks

Material of Filter Tank Construction

The stainless steel tank option is the most expensive, but is recommended due to the increased flexibfity provided by metal-tank filters and the long-term reduced maintenance associated with stainless steel. The metal tank options offer greater quality control because most of the work is done in the manufacturer's plant and does not depend upon field work by the Contractor to place and form concrete to exact dimensions. Also, stainless steel does not require recoating to preserve its integrity as does painted steel tank systems. The metal tank systems minimize field labor to reduce installation costs.

Cost ($)

Any of the three options is acceptable, and would provide equivalent process performance.

6.8 Effluent Filter Arrangement and Support

The effluent filters will be located in the proposed Filter UV Disinfection Building, shown in Figure

6.8.1. The building is proposed to be a pre-engineered structure 50 feet by 55 feet in size selected to fit into the overall site architecture. Figure 6.8.2 shows the location of the Filter/UV Disinfection Burlding on the plant site.

The three filter tanks wlll be set on top of the operating floor and bolted down for seismic reasons.

The tanks will be moveable if this is required in the future. The flow to the filters will be by gravity from the new metal removal flocculator/clarifiers. The influent flow would be split between the tanks using a manually adjusted valve in each of the inlet pipes and the inlet weirs in each filter tank.

Relative flow balancing is required, but absolute accuracy is not critical to performance.

Filtered water overflows from each Wter tank by gravity and flows to the UV system in a 27-inch pipe. Backwash water and tank drain water will collect in a sump and be pumped back to the plant

I?\-I~mjects\l2G517 - ~OMPOC\\X'P\~roject~Report\1033979 Preliminary Desip Report Finddoc February 2005

Effluent Filtration (TM 4) 6-10 headworks for reprocessing. A Panhall flume will be provided downstream of the W dxinfection system to provide plant flow measurement. The flow can be diverted to the emergency storage basin at the effluent pump station and returned to the activated sludge system for reprocessing if necessary.

The building d be vendated, but not climate conditioned. A separate electrical room will be provided for filter controls and panels in conjunction with W system control panels. The control panels will be tied into the plant SCADA system to facilitate remote monitoring and control.

Channel Drain Level Contml

,- Me\ Gales [;;I"," / weirs

Electrical Control Room

PLAN

SECTION

Figure 6.8.1 Effluent Filters and W Disinfection Building Layout

Effluent Filtration (TM 4) 6-1 1

Figure 6.8.2 Location of the Filter/UV Disinfection Building at Site

P:\-lhjrcts\l26517 - LOhTPOC\\~1'\PmjrctfReport\Il133979 Preliminary Dcsip Report Final.Joc February 2005

Effluent Filtration ('I'M 4) 6-1 2

Preliminary Construction Costs

The preliminary estimate of probable construction cost for effluent filters is $2,598,287 as indicated for TM 04 in Table 2.1-2. A detailed breakdown of this estimate is provided in Appendtx B.

6.10 Conclusions and Recommendations

The following conclusions and recommendations are made relative to effluent filtration for the

LRWRP.

A substantial number of filter options are available for consideration and have been tested and approved for Title 22 applications in California.

The recommended filter type is the cloth media disc filter. This filter has excellent performance characteristics for plant effluent, has been tested and approved by DHS for Title 22 applications, and is relatively easy to operate and maintain.

Three filter tanks are recommended with two for service and one for required redundancy.

Stainless steel tanks are recommended due to ease of installation and minimal long-term maintenance. The filter tanks will be moveable if required in the future.

The stainless steel tank option provides greater quality control because it is factory assembled and does not require field labor for complex concrete work.

Filter influent will be equalized ahead of filtration so that the design flow is 5.5 mgd

The filters will be installed in the proposed Filter/UV Disinfection Building, as shown in Figure 6.8.1. The buildmg will be of a pre-engineered structure approximately 50 feet by 55 feet in size.

The filter hydraulic loading rate will be 2.96 gpm/ft2 at the design flow rate of 5.5 mgd with one unit out of service. This meets DHS limitations for Title 22 applications.

A monorail or bridge crane arrangement (to be determined during final design) wdl be installed over each filter to assist with infrequent component lifts.

I>:\-Prr~jects\l26517 - LO~~OC\\~P\l'r~~jrct-K~1port\I033979 Prrliminaq Design Report Final.doc February 2005

ATTACHMENT A

CLOTH MEDIA DISC FILTER INSTALLATIONS

AND RELATED INFORMATION

AQUA-AEROBIC SYSTEMS, INC.

AquaADFB Filter Reference List

(With UV Disinfection) 1 1/23/04

18 MlLE CREEK MID. REGIONAL (AKA PICKENS COUNTY) Project ID: 105391 and 105391A 967 Willie Wonka Drive - Off M&M Road Central, South Carolina 29630 8641306-7600 Eddie Barker, Operations Manager Start-up Date: January and June, 2000 Waste Characteristics: MunicipallDomestic Treatment Objective: TSS Avg. Daily Flow: 0.929 mgd

1 Tertiary Filters 3 ADF Disk Package Filters

18 MlLE CREEK UPPER REGIONAL (AKA PICKENS COUNTY) Project ID: 105610 151 Clearwater Drive Liberty, South Carolina 29657 8641306-7600 Eddie Barker, Operations Manager Start-up Date: June, 2000 Waste Characteristics: Municipal/Domestic Treatment Objective: TSS Avg. Daily Flow: 0.929 mgd

1 Tertiary Filter 1 ADF Disk Package Filter

APPOMATTOX WATER RECLAMATION FACILITY

Project ID: 105671 AIB Ethel Street Appomattox, Virginia 24522 4341352-7659 Charles Garrett, Chief Operator Start-up Date: June, 2001 Waste Characteristics: MunicipallDomestic

Treatment Objective: Nitrification Avg. Daily Flow: 0.315 mgd

1 AquaSBR Basins 2 AquaDDM Mixers 2 Pumps 2 6 x 4 Gravity Decanters 1 PLCJSCADA Controls 2 Influent Baffles 6 Retrievable Coarse Bubble Diffusers 3 Positive Displacement Blowers

Post Equalization Basin 3 Pumps 1 Positive Displacement Blower 4 Fixed Coarse Bubble Diffusers

Aerobic Digester Basins 2 Positive Displacement Blowers 2 Fixed Coarse Bubble Diffusers 2 Pumps

Tertiary Filters 2 ADF Disk Package Filters

COLDWATER WWTP EXPANSION

Project ID: 105700 100 Jay Street Coldwater, Michigan 49036-0469 51 71279-9531 David L. (Woody) Woodman, WWTP Superintendent Start-up Date: April, 2001 Waste Ch,aracteristics: Municipal/Domestic Treatment Objective: TSS Avg. Daily Flow: 3 mgd

Tertiary Filters 3 ADF Disk SS Package Filters

HAILEY WASTEWATER TREATMENT PLANT

Project ID: 100190A 4297 Glenbrook Drive Hailey, Idaho 83333 2081578-221 1, ext. 10 Roger Parker, Superintendent Start-up Date: April, 2000 Waste Characteristics: MunicipallDomestic Treatment Objective: TSS Avg. Daily Flow: 1.39 mgd

Tertiary Filter 1 ADF Disk Concrete Filter

KEARNY WWTP

Project ID: 102573A 501 Veterans Ave.

Kearny, Arizona 85237 5201363-0209 Raymon Carnacho, Public Works Manager Start-up Date: May, 2000 Waste Characteristics: MunicipallDomestic Treatment Objective: TSS Avg. Daily Flow: 0.1 75 mgd

Tertiary Filter 1 ADF Disk Concrete Filter\

KITTITAS WWTP

Project l D: I05900 600 Industrial Loop Kittitas, Washington 98934 5091968-0225 Cecil Howard, WWTP Operator and Water Distribution Manager Start-up Date: September, 2001 Waste Characteristics: MunicipallDomestic Treatment Objective: TSS Avg. Daily Flow: 0.25 mgd

Tertiary Filter 1 ADF Disk Package Filter

LICKING WWTP

Project ID: 105708A 225 N. Highway 63 Licking, MO 65542 5731674-3403 Alan Quick Start-up Date: September, 2000 Waste Characteristics: Municipal1Domestic Treatment Objective: TSS Ave. Daily Flow: 0.3 mgd

Tertiary Filter 2 ADF Disk Package Filters

PALM VALLEY WRF

Project ID: 106039 4620 E. Elwood Street, #B14 Phoenix, Arizona 85040-1 961 4801557-8525 Rick Rhoades, PERC Operations Manager Start Up Date: March 7,2002 Waste Characteristics: MunicipalIDomestic Treatment Objective: TSS I NTU 1 Reuse Avg. Daily Flow: 4.1

Tertiary Filter(s) 3 ADF Disk Package Filters

PECAN BRANCH WWTP (CITY OF GEORGETOWN)

Project l D: 1 06063

3502 FM971

Georgetown, TX 78626 51 21930-3538 (OMI, contract operator) Randy Massie (OMI, contract operator) Start-up: August, 2002 Waste: MunicipallDomestic Treatment: TSS Ave. Daily Flow: 1.5 mgd

1 ADF Disk Concrete Filter

SIMMONS FOODS

Project ID: 100494E Rte. 1, Box 121 Southwest, Missouri 64863 41 71762-3271 Andy Brashear Start-up Date: December, 1998 Waste Characteristics: IndustriallPoultry Treatment Objective: TSS Avg. Daily Flow: 1.5 mgd

1 Tertiary Filters 2 ADF Disk Package Filter

SPIRIT MOUNTAIN CASINO

Project ID: 1051 88 2700 SW Salmon River Hwy (1 8) Grand Ronde, Oregon 97347 8001760-7977, ext. 3995 Robert Jones, Operator Start-up Date: June, 1998 Waste Characteristics: Municipal/Resort Treatment Objective: TSSlReuse Avg. Daily Flow: 0.464 mgd

Tertiary Filters 2 ADF Disk Package Filters

WASHINGTON STATE DEPT. OF CORRECTIONS (AKA

SHELTON)

Project ID: 1 O2983B West 2321 Dayton Airport Road Shelton, WA 98584 3601426-4433, ext. 5779 George Campbell Start-up Date: September, 1997 Waste Characteristics: Municipal/Domestic Treatment Objective: TSS Avg. Daily Flow: 0.2 mgd

I ADF Disk Package Filter

ATTACHMENT B

INTEGRATED APPROACH FOR DESIGN AND OPERATION OF WATER

RECLAMATION FACILITIES

B. Narayanan, K. Bourgeous, A. Salveson Carollo Engineers, Walnut Creek, CA

Introduction

Water reclamation facility design has traditionally been based on an "add-on" approach, wherein the facility is viewed as a combination of two distinct plants - a secondary treatment plant at the front end and an advanced treatment plant at the back end. The secondary treatment plant is typically designed and operated as a wastewater plant for removal of conventional pollutants, and all of the "reclamation" occurs at the back end in the advanced treatment (filtration / disinfection) processes. While the origin of this philosophy is understandable in that most reclamation facilities were expansions of existing secondary treatment facilities, the approach results in higher capital and operations costs because it fails to utilize the synergies that exist between the different treatment processes. A better approach is to view all processes as part of a single integrated system with the common end goal of producing high quality reclaimed water. This paper discusses the merits of such an integrated approach based on the linkages between the different processes in a water reclamation facility.

Traditional Approach to Water Reclamation Facility Design

Water reclamation facilities have typically been expansions of existing secondary treatment wastewater treatment plants. As a result, there has been a tendency to view the overall facility as a combination of two distinct plants - a secondary treatment wastewater treatment plant at the front end and an advanced treatment plant at the back end. All of the "reclamation" is assumed to occur at the back end in the advanced treatment (filtration / disinfection) processes, with little or no connection to the secondary treatment process. Even within the "reclamation" plant, there is the tendency to view each component process as a separate identity responsible for a separate and distinct function. Thus, for example, the role of the filtration system is often seen as limited to turbidity reduction, with the disinfection system being responsible for pathogen inactivation. This traditional approach fails to adequately consider the linkages that exist between the different treatment processes in a water reclamation facility.

Linkages Between Different Treatment Processes

1 . Linkages Between Secondarv Treatment O~eration and Filtration Performance

Recent studies have documented the impact of influent particle size distribution (PSD) on filter performance (Bourgeous et al., 2003). It has been shown that most filter systems do not effect much removal of particles less than 10 pm in size (Figure 1). Consequently, if a high number of these smaller-sized particles are present, a coagulation / flocculation step is often needed upstream of the filtration process. While this approach does effectively reduce turbidity in the filter effluent, it is also expensive, both from a capital cost and operational cost standpoint.

100000

I Secondary Effluent (1 7 NTU) - E r 10000

(II

Filtered Effluent (no chemical coagulation) (12 NTU) c (U A Coagulated I Flocculated I Filtered Effluent (1 NTU) N .- : -1 1000 2i.E rn a E - .O a r 0 100 fi u-k n 10 S z

Particle diameter, pm

Figure 1. Effect of Coagulation 1 Flocculation on Removal of Particles through Filtration

Recognizing the linkages between the secondary treatment process and filtration opens up the potential for alternative solutions. Recent work (Bourgeous et al., 2003) has shown that the type and mode of operation of the secondary treatment process can have a significant impact on the secondary effluent (filter influent) PSD (Figure 2).

gActivated Sludge (High SRT)

Particle diameter, pn

Figure 2. Effect of Type of Secondary Treatment Process on Secondary Effluent PSD

As shown in Figure 2, increasing the solids retention time (SRT) of the activated sludge (AS) process I significantly reduces the number of smaller-sized particles, to the extent where a coagulation / flocculation step preceding filtration is no longer necessary. Thus an increase in the SRT of the activated sludge process could be a more cost-effective alternative to the construction of separate coagulation / flocculation facilities upstream of the filters.

2. Linkages Between Secondary Treatment Operation and Disinfection

The type of secondary treatment process also appears to have an effect on the design and performance of disinfection systems. For UV disinfection systems, the UV transmittance at 254 nanometers (UVT254) of the wastewater is an important consideration that impacts both on the size (i.e., the number of lamps) of the UV disinfection system and its performance. Figure 3 shows measurements of effluent UVT254 collected from different operating treatment plants, sorted by type of secondary treatment process.

Activated Sludge Activated Sludge (High SRT)

80 s 70 6

Solids Contact

!= 60 m Y

V)

!= 40 E

3 20

Figure 3. Effect of Type of Secondary Treatment Process on Secondary Effluent UVT254

As can be seen, the UVT254 varies considerably depending on the type of secondary treatment process utilized. In general, the secondary effluents from activated sludge processes show a higher UVT compared to the effluents from trickling filter or trickling filter / solids contact processes. Within the activated sludge processes, the UVT appears to increase with increasing SRT. Therefore, the type and mode of operation of the secondary treatment process can impact the performance of the disinfection system.

3. Linkages Between Filtration and Disinfection

Turbidity reduction has traditionally been viewed as the main function of the filtration system in water reclamation facilities. However, an equally important function is to condition the water for disinfection,

- specifically with respect to the removal of particles that could potentially shield coliform bacteria from 1 the disinfectant. Previous research (Schieble, 1987; Parker and Darby, 1995) has shown that coliform bacteria can associate with particles to such a degree that they can become completely shielded from UV light, resulting in a residual coliform concentration. A particle size of around 10 pm has been hypothesized to be the minimum size needed to cause this shielding effect. Therefore, the number of particles larger than 10 pm that are present in an effluent can significantly affect the performance of a UV disinfection system. Viewed from this perspective, it is clear that one of the main goals of the filtration system should be to maximize the removal of these particles greater than 10 pm. This is especially important given some recent work that shows that the turbidity measurement does not accurately capture the presence of particles greater than 10 pm. In fact, as shown below in Figure 4, it is possible for two water samples with similarly low turbidity measurements to have significantly different concentrations of particles greater than 10 pm, resulting in different levels of disinfection performance.

Figure 4 compares the side-by-side performance of two full-scale filters.

Filter A (1.6 NTU, 70% transmittance) Y T d a l Coliform = 790 MPN 1 100 mL at UV dose of 58 r n ~ l c m ~

Filter B (1.5 NTU, 69% transmittance) .. Total Coliform = 10 MPN 1 100 mL at

UV dose of 58 m~lcm' . I

Particle diameter, prn

Pigure 4. Impact of Large Particle Removal through Filtration on UV Disinfection Performance

Even though the filtered effluents had similar turbidities (1.5 NTU and 1.6 NTU) and similar transmittance values, there was a significant difference in the concentrations of particles greater than 10

pm. This resulted in a significant difference in the level of disinfection achieved when the effluent samples from both filters were exposed to 58 m ~ i c m ~ of UV light from a collimated beam apparatus.

The filter effluent with more larger particles had a higher total coliform bacteria concentration (790 MPN1100 mL) after disinfection relative to the filter effluent with fewer larger particles (10 MPN/100

The recognition of this linkage between filtration and disinfection allows for a more rational assessment of filtration systems, both in the selection of filters during design, as well as for diagnostic evaluations of filter performance. , Conclusions

Based on the discussion presented in this paper, it is clear that there are significant linkages between the different processes in a water reclamation facility. The traditional approach to the design and operation of reclamation facilities does not adequately consider these linkages, and thereby fails to utilize potential synergies between the processes. A more integrated approach to facility design and operation, one that considers the impact of these inter-linkages between the different treatment processes, can offer several benefits, including potential reductions in construction and operating costs.

References

Bourgeous, K., Narayanan, B., and G. Deis (2003), "Particle Size Distribution Testing as a Tool for Optimizing Filtration and Disinfeetion Systems," Proceedings, WEFTEC 2003, Anaheim, CA

Schielble, 0. (1 987), "Development of a Rationally-Based Design Protocol for the Ultraviolet Light Disinfection Process," Journal Water Pollution Control Federation, 59:25.

Parker, J., and J. Darby. (1995), "Particle-Associated Coliforrn in Secondary Effluents: Shielding and Ultraviolet Light Disinfection," Water Environment Research, 67: 1065.