8
1 JUNE 2005 W ith the development of total maximum daily loads for nutri- ent discharges, many wastewater treatment plants (WWTPs) face stringent nutrient limits. Deep-bed denitrifica- tion filters that remove nitrogen and solids are a proven technology for treating wastewater to meet low total nitrogen (TN) limits. A variety of denitrification filters are available, offering differ- ent features and levels of experience, and WWTP managers may have a hard time selecting the sys- tem that best meets their needs. To simplify the evaluation process, the following comparison of denitrification filter equipment and performance is intended to highlight some of the similarities and differences among the systems. Filter Configurations A deep-bed filter capable of concurrent de- nitrification and solids removal was first pat- Evaluating Denitrification Filters Before selecting a denitrification filter, designers should compare all available systems and assess their performance to date Christine deBarbadillo, Robert Rectanus, Shannon Lambert, David Parker, Jeff Wells, and Robert Willet USFILTER ©2005 Water Environment Federation. All Rights Reserved. For website posting only. Bulk printing prohibited.

Evaluating Denitrification Filters - Environmental Expert discharges, many wastewater treatment plants (WWTPs) face stringent nutrient limits. ... (5 scfm/ft2) Continuous through air

  • Upload
    vutruc

  • View
    213

  • Download
    0

Embed Size (px)

Citation preview

1J U N E 2 0 0 5

W ith the development of total

maximum daily loads for nutri-

ent discharges, many wastewater

treatment plants (WWTPs) face

stringent nutrient limits. Deep-bed denitrifica-

tion filters that remove nitrogen and solids are

a proven technology for treating wastewater to

meet low total nitrogen (TN) limits. A variety of

denitrification filters are available, offering differ-

ent features and levels of experience, and WWTP

managers may have a hard time selecting the sys-

tem that best meets their needs. To simplify the

evaluation process, the following comparison of

denitrification filter equipment and performance

is intended to highlight some of the similarities

and differences among the systems.

Filter ConfigurationsA deep-bed filter capable of concurrent de-

nitrification and solids removal was first pat-

Evaluating Denitrification Filters

Before selecting a

denitrification filter,

designers should

compare all

available systems

and assess their

performance to date

Christine deBarbadillo,

Robert Rectanus, Shannon Lambert,

David Parker, Jeff Wells, and

Robert Willet

US

FIL

TER

©2005 Water Environment Federation. All Rights Reserved. For website posting only. Bulk printing prohibited.

2 W E & T

Downflow denitrifi-cation filters at the city of Largo, Fla.

SE

VE

RN

TR

EN

T S

ER

VIC

ES

ented in 1973. Dravo Corp. — now known as

Dravo Lime Co. (Pittsburgh) — and later Tetra

Technologies (The Woodlands, Texas) pioneered

the development of denitrification filter technol-

ogy. Following the expiration of the original pro-

cess patents, several other filter manufacturers

have developed a denitrification option for their

equipment. Moreover, some utilities in the United

States and Europe have retrofitted existing filters

to achieve denitrification.

Two main denitrification filter configurations

— downflow filters and upflow continuous-

backwash filters — are available. With both

configurations, methanol (or another readily

biodegradable carbon source) is added to waste-

water ahead of the filter to enable denitrifying

bacteria to grow.

Downflow denitrification filters operate in

a conventional filtration mode and consist of

gravel and sand supported by an underdrain (see

photo, above). Manufacturers include Severn

Trent Services (Fort Washington, Pa.), maker of

the TETRA® Denite® system; F.B. Leopold Co. Inc.

(Zelienople, Pa.), maker of the elimi-NITE system;

and USFilter Davco Products (Thomasville, Ga.),

maker of the Davco denitrification filter.

Wastewater enters a downflow filter over

weirs located along the length of the filter bed

on both sides. Filter effluent is conveyed from

the bottom of the filter over a control weir into a

clearwell. Filters must be taken out of service at

regular intervals for a short backwashing cycle

consisting of air scouring and backwashing with

air and water. Nitrogen-release cycles are needed

to remove nitrogen gas bubbles that are pro-

duced during denitrification and accumulate in

the media. To avoid sending slugs of spent back-

wash water to a plant’s headworks, a mudwell

normally is provided for equalization. The piping

for the filter influent and backwash is similar to

that of conventional filters and can be housed in

an indoor pipe gallery or installed outdoors.

Upflow continuous-backwash filters differ in

that influent wastewater flows upward through

the filter countercurrent to the movement of the

sand bed. The filters are supplied as modular

units installed in steel tanks or in multiple cells in

concrete basins (see photo, p. 26). Manufacturers

include Parkson Corp. (Fort Lauderdale, Fla.),

maker of the DynaSand filter, and Paques bv

(Balk, Netherlands), maker of the Astrasand filter.

Since October 2003, USFilter Davco Products has

had a license agreement with Paques to supply

this filter in the United States and Canada.

Wastewater enters an upflow continuous-back-

wash filter at the top and is conveyed downward

through the feed pipe and distributed to the filter

bed through feed radials (see Figure 1, below).

3J U N E 2 0 0 5

After traveling upward

through the media, ef-

fluent wastewater is

removed at the top of

the filter. Sand media

slowly travel down-

ward and are drawn

into an airlift pipe in

the center of the filter.

Compressed air is in-

troduced to the airlift,

drawing sand upward

and scouring it. At the

top of the airlift, the

media are returned

to the filter bed. Filtered water rises through a

separator that removes lighter dirt particles by

washing them away and returns the large, heavy

sand grains to the top of the filter bed. The reject,

or backwash, water continuously exits near the

top of the filter. The reject-water weir is set at a

lower elevation than the effluent weir to enable

clean water to enter the washer and separator

continuously by differential head, eliminating

the need for typical backwash-supply pumps.

Individual filter cells are not taken out of service

for backwashing, enabling a relatively simple pip-

ing and valve arrangement.

Filter ManufacturersWhen designing a denitrification filter, one

must examine differences in equipment and ex-

perience offered by the manufacturers (see Table

1, p. 27). Major design considerations include a

manufacturer’s experience, system performance,

and such factors as design loading rates, influent

weir, media, underdrain, process control, and

methanol feed control.

With more than 25 years of full-scale experi-

ence, the TETRA® Denite® system has been

installed in greater numbers than any other

denitrification filter. Leopold has manufactured

conventional water and wastewater filters for

several years and recently began offering the

elimi-NITE filter, an adaptation of its conventional

filter, for tertiary denitrification applications.

USFilter Davco has offered denitrification filters

for about 15 years, mostly at small installations

with capacities less than 3800 m3/d (1 mgd).

Although use of DynaSand filters is not wide-

spread, several installations capable of denitri-

fication have been constructed in the United

States and Puerto Rico within the past 15 years.

Five Astrasand installations in Europe currently

are operated for denitrification, with the first

one commissioned in 1999. The first Astrasand

unit in the United States is under construction

at the U.S. Army’s Aberdeen Proving Ground in

Maryland.

All the filter manufacturers indicate that

their equipment can reduce concentrations of

nitrate–nitrogen and nitrite–nitrogen in effluent

to less than 1 mg/L. The TETRA Denite system

is guaranteed to meet this target, and its perfor-

mance is well documented, with full-scale data

from multiple facilities. Recent full-scale data

also show that the system can achieve an efflu-

ent nitrate–nitrogen concentration of less than

0.5 mg/L under cold weather conditions. Davco

filters have been installed at some small WWTPs

in Florida that have met an effluent limit of

3 mg/L of TN. Although the elimi-NITE system has

limited full-scale experience in the denitrifica-

tion mode, data collected at two installations in

North Carolina show that concentrations of less

than 1 mg/L nitrate–nitrogen can be met at low

loading rates.

Full-scale data from several DynaSand facilities

in Puerto Rico show that the filters can achieve

concentrations of less than 1 mg/L nitrate–

nitrogen and nitrite–nitrogen. This information

is supplemented with data from pilot test-

ing conducted in 1989 by the University of

Concrete installa-tion of Astrasand® filter modules.

Figure 1. Schematic of Dynasand® Upflow Continuous Backwash Filter

PAQ

UE

S B

V/U

S F

ILTE

R

PAR

KS

ON

CO

RP.

4 W E & T

Florida (Gainesville) and testing completed in 2005 at the

Hagerstown (Md.) WWTP. Extensive full-scale data from

the De Groot Lucht Sewage Treatment Plant (Vlaardingen,

Netherlands) show that the Astrasand filters reduced aver-

age influent nitrate–nitrogen concentrations of 18 mg/L to

about 2 mg/L at cold wastewater temperatures.

All the filters can achieve effluent concentrations of 5

mg/L or less of total suspended solids (TSS). In addition,

the TETRA® Denite® system has received conditional

acceptance from the California Department of Health

Services, under its so-called Title 22 reuse regulations, for

use as a standard filtration system and as a denitrifica-

tion filter. The elimi-NITE, Davco, and DynaSand systems

have received conditional acceptance for filtration only

Table 1. Comparison of Denitrification Filter Manufacturers and Equipment

Manufacturer/filter

Severn Trent Services/

TETRA® Denite®

F. B. Leopold/ elimi-NITE

USFilter/Davco Parkson/DynaSand

Paques and USFilter/Astrasand

Flow regime Downflow Downflow Downflow Upflow Upflow

Underdrain T-block; concrete-filled, HDPE jacket

Universal Type S

HDPE blockPipe lateral; or

Multiblock HDPE block

None required None required

Air header arrangement

SS box header; laterals beneath

underdrain

SS header across filter; laterals

SS air header; 50-mm (2-in.) laterals

Vertical air lift Vertical air lift

Media 457 mm (18 in.) graded gravel,

1.8 m (6 ft) of 6 × 9 mesh silica sand, uniformity coefficient 1.35,

0.8 minimum sphericity

381 mm (15 in.) graded gravel,

1.8 m (6 ft) of 6 × 12 mesh sand

2 layers support gravel,

1.8 m (6 ft) of 6 × 9 mesh sand

1.35 to 1.45 mm subround media or 1.55 to 1.65 mm subangular media with uniformity

coefficient of 1.3 to 1.6; 2-m (6.6-

ft) bed depth

1.2 to 1.4 mm sand, 2-m (6.6-ft)

bed depth

Nitrogen-release cycle

Initiated by headloss or time-controlled cycle;

Speed Bump controls

Initiated by headloss or time-controlled cycle

Initiated by headloss or time-controlled cycle

None required None required

Backwash water and air requirement

244 L/min·m2 (6 gal/min·ft2);1.5 m3/min·m2

(5 scfm/ft2)

244 L/min·m2 (6 gal/min·ft2);1.5 m3/min·m2

(5 scfm/ft2)

407 L/min·m2 (10 gal/min·ft2);1.5 m3/min·m2

(5 scfm/ft2)

Continuous through air lift

and sand washer

Continuous through air lift and

sand washer

Influent weir type Curvilinear weir block

Curved stainless steel weir

Varies Feed radials at bottom of unit

Feed radials at bottom of unit

Backwash flow as percent of forward flow

<5; often 1 to 2 2 Not documented 3 to 5 3 to 12

Patented features

T block underdrain, curvilinear weir block, Speed

Bump, TetraPace, TetraFlex

Universal

underdrain and features

None None None in United States; Astracontrol in

Europe

HDPE = high-density polyethylene.

SS = Stainless steel

5J U N E 2 0 0 5

and have not been tested in denitrification mode. The

Astrasand filter has not undergone testing to receive

conditional acceptance by the Department of Health

Services.

Although tertiary denitrification filters historically have

been used for nitrogen removal only, the need to meet

stringent effluent total phosphorus (TP) limits at the same

time recently has become an issue in some locations. For

example, the State of Maryland plans to require WWTPs to

upgrade to meet effluent TN and TP limits of 3 and 0.3 mg/

L, respectively, to reduce nutrient loads to the Chesapeake

Bay. Although many plants in Florida and elsewhere meet

moderate TP limits of about 1 mg/L while complying with

stringent TN limits, few facilities have experience operat-

ing denitrification filters to produce low effluent TN and

TP concentrations simultaneously. As the TP limit decreas-

es, phosphorus limitations on the denitrification process

become a concern: Enough phosphorus must be present

to meet the requirements for growth of the denitrifying

bacteria, but the phosphorus concentration must remain

low enough to meet the effluent limit reliably.

Several WWTPs using TETRA ® Denite ® f i l -

ters have achieved low ef f luent phosphorus

concentrations while producing eff luent con-

centrations of less than 1 mg/L nitrate–nitrogen and nitrite–

nitrogen. A key component of the pilot test of the

DynaSand filters at the Hagerstown WWTP involved add-

ing chemicals to the denitrification filter influent to aid in

phosphorus removal. The pilot results show that adding

ferric chloride enabled the system to reduce influent TP

levels of approximately 1 mg/L to less than 0.3 mg/L in the

effluent while also reducing concentrations of nitrate–ni-

trogen and nitrite–nitrogen from approximately 7 to less

than 1 mg/L. Automatic chemical-dosing control and on-

line nutrient monitoring are important components when

operating in this mode.

Design Loading RatesAs reported in the literature, design procedures for

denitrification filters have focused on hydraulic and mass

loading guidelines. Based on early data, design curves

were developed to compare nitrate-removal efficiency

versus empty-bed detention time for downflow denitrifi-

cation filters, and these findings have been included in

subsequent textbook publications.

Designers historically have depended on filter manu-

facturers to provide design guidance. However, differing

levels of operating experience among manufacturers and

the wide range of design loading rates cited in the litera-

ture often prompt questions regarding appropriate loading

rates for denitrification. The industry would benefit over-

all by establishing better design criteria for achieving low

nitrate–nitrogen concentrations, particularly in temperate

and cold climates. A summary of design loading rates from

different sources is given in Table 2 (see p. 6).

Filter Influent WeirsDownflow denitrification filters are operated at a vari-

able level and generally have a significant drop over the

influent weir, resulting in entrainment of dissolved oxygen

(DO). The increase in DO reduces the efficiency with

which the filter removes nitrate and increases methanol

consumption. To mitigate this problem, the TETRA Denite

system is provided with a patented curvilinear weir block

to encourage laminar flow down the wall to minimize DO

entrainment. The elimi-NITE system can be installed with

a curved stainless steel weir. Leopold also has suggested

that operating the system in a constant-level mode would

reduce the elevation drop from the influent weir, thereby

decreasing the level of DO entrainment. Because influent

in upflow continuous-backwash filters is conveyed to

the feed radials within the filter bed through submerged

manifold piping, DO entrainment over the influent weir is

less of an issue for those filters.

MediaTable 1 (p. 27) lists the preferred media of each filter

manufacturer. The 6 × 9 mesh media used in the Tetra Denite

system meet relatively stringent standards for uniformity

and sphericity. The uniform and relatively spherical media

reportedly allow for more rolling and contact with other

media grains, resulting in more effective backwash and ni-

trogen-release cycles and, ultimately, lower backwash water

volume requirements. Davco filters can be supplied with

the same media. The elimi-NITE filter typically is supplied

with 6 × 12 mesh media, but the 6 × 9 mesh media can be

provided if desired. Finer media are used with the DynaSand

and Astrasand filters.

UnderdrainEarly experience with downflow denitrification filters

suggested that nozzle underdrains were prone to foul-

ing and failure. To avoid this issue, the manufacturers

have developed unique block underdrains (see Figure

2). Severn Trent Services offers the TETRA® T-block un-

derdrain, which is specifically designed for bioreactor

service and consists of concrete-filled blocks enclosed

in high-density polyethylene (HDPE). Leopold developed

its Universal Type S underdrain, which consists of HDPE

blocks. Although existing Davco filters were constructed

with pipe lateral underdrains, new installations would be

supplied with the Multiblock HDPE underdrain. Experience

is limited with the underdrains used with the elimi-NITE

and Davco systems, and it is unclear whether fouling will

be a significant issue over the long term when operating

in the denitrification mode at moderate to high loading

rates. Upflow continuous-backwash filters do not require

an underdrain.

Nitrogen-Release CycleThe TETRA Denite system offers a patented nitrogen-

release cycle control package, known as SpeedBump,

6 W E & T

that pumps backwash water up through the filter for 30

seconds to 2 minutes. The influent valve to the filter re-

mains open to minimize filter downtime. The elimi-NITE

and Davco systems offer nitrogen-release cycles that fully

close the influent valve, and the additional time required

for the nitrogen-release cycle should be accounted for in

the filter design. Because the DynaSand and Astrasand

upflow systems operate in the same direction that the ni-

trogen gas travels, and the gas also is drawn into the airlift,

a separate degassing cycle is unnecessary.

Backwashing and Filter ControlsDuring operation of the denitrification filter, solids

removed from the wastewater accumulate in the media,

and additional solids are formed from the growth of de-

nitrifying bacteria. To clean the media, backwashing cycles

for the downflow filters are initiated based on increased

head loss through the filter or on a timed basis. All three

manufacturers of downflow filters offer air scouring and

air–water backwash as part of the backwash cycle.

The TETRA® Denite®, elimi-NITE, and Davco filtration

systems offer integrated control packages for backwash-

ing, air-scour, and nitrogen-release cycles. For installations

in which only partial denitrification is required, the TETRA

Denite system is available with another patented control

system that allows some filter cells to be operated for full

denitrification, while others are operated in parallel for

TSS removal only at a higher hydraulic throughput.

The DynaSand and Astrasand systems operate with a

small continuous-backwash stream. Media bed turnover

rates historically have ranged from 305 to 457 mm/h (12 to

18 in./h) or four to six bed turnovers per day for conven-

Table 2. Summary of Design Guidance for Denitrification Filters

Source

Hydraulic loading rate [L/min·m2 (gal/

min·ft2)]

Volumetric mass loading rate [kg NO3-N per m3/d

(lb NO3-N per ft3/d)] Other information

Manual: Nitrogen Control (U.S. Environmental Protection Agency, 1993)

41 to 82 (1 to 2), 30 minutes empty bed contact time

0.29 to 1.6 (0.018 to 0.1)

Design curves from Savage, E.S. (1983), “Biological Denitrification Deep Bed Filters,” presented at the Filtech Conference, Filtration Society, London, England.

1.33 (0.083) [referenced Tetra data]

Hydraulic loading rate versus effluent nitrate–nitrogen concentration (referenced Tetra data)

Biological and Chemical Systems for Nutrient Removal, Special Publication (Water Environment Federation, 1998)

0.24 to 3.2 (0.015 to 0.2) depending on tem-perature

Design curves from Savage, 1983

Wastewater Engineering, Treatment and Reuse (Metcalf & Eddy, 2003)

41 to 82 (1 to 2) at 20oC

1.4 to 1.8 (0.087 to 0.112) at 20oC 20 to 30 minute empty bed

contact time20 to 61 (0.5 to 1.5) at 10oC

0.8 to 1.2 (0.05 to 0.075) at 10oC

Severn Trent Services TETRA®Denite®

<123 (<3) at aver-age flow; <308 (<7.5) peak hydrau-lic with one cell out of service

Determined using pro-cess model

In-house kinetic model, extensive full-scale data

F.B. Leopold 41 to 82 (1 to 2) 1.12 (0.07) Full-scale data (North Carolina)

USFilter/Davco 82 (2) Not available Full-scale data (Florida)

Parkson Up to 183 (4.5) 0.24 to 1.9 (0.015 to 0.12)

Full-scale and pilot data (Puerto Rico, Maryland)

Paques/USFilter 168 (4.1) 2.08 (0.13) Full-scale data (Netherlands), dry weather flow only

7J U N E 2 0 0 5

tional filtration. Recent pilot testing of the DynaSand filter

indicated that bed turnover rates of 203 to 254 mm/h (8

to 10 in./h) were effective for denitrification. As a process

monitoring tool for the Astrasand filter, the Astrameter

system is used to measure the sand circulation rates at

several locations throughout the filter.

Questions remain regarding the bed turnover rate

(backwash frequency) and how it relates to maintaining

good solids removal while supporting sufficient biomass

for denitrification. Available for use with the Astrasand

filter, the Astracontrol system was developed to maintain

biological activity within the filter under varying condi-

tions. The control system continuously adjusts the media

movement and washing rate to maintain a fixed volume

of active biomass in the filter. Parkson has indicated that

it may be necessary to change the bed turnover rate in

the DynaSand system to meet a specific requirement.

However, the company has not seen a need to adjust it

during routine operation.

Methanol Control SystemMethanol normally is dosed to the filter influent before

it is divided among filter cells. A patented methanol-con-

trol system named TetraPace is available for use with the

TETRA® Denite® system. TetraPace dispenses methanol

based on the filter influent flow rate and the concentra-

tions of nitrate in the influent and effluent, as measured

by an online nutrient analyzer. When this control system

is used with the TETRA Denite system, the manufacturer

guarantees no net increase in total organic carbon across

the filter.

The other manufacturers suggest using the filter influ-

ent flow rate and nitrate concentration to determine the

methanol dosage via either a flow-paced or feed-forward

automatic control system. Although a feed-forward control

scheme can reasonably match methanol dosing to actual

requirements, it may be difficult to avoid periods of slight

overdosing and the resulting increase in concentrations of

biochemical oxygen demand (BOD) in the filter effluent.

In cases in which effluent BOD and nitrate–nitrogen limits

are less stringent, the need for a high level of methanol

control is related to optimizing chemical usage. However,

tighter measures for controlling methanol, such as that

offered by the TetraPace system, can provide significant

advantages to a facility that has a stringent BOD limit of

5 mg/L or lower and must achieve low (less than 1 mg/L)

concentrations of nitrate–nitrogen and nitrite–nitrogen.

Cost and Bidding ConsiderationsWhen selecting a denitrification filter, wastewater

professionals should consider several factors related to a

system’s capital costs. The most critical factor involves the

overall area called for in the design. Depending on the appli-

cation and overall effluent requirements, it may be desirable

at times to use a more conservative design for filters with

less full-scale operating experience in meeting the required

limit. Alternately, pilot testing can be conducted to verify

the design loadings. Because they need backwash pumps

and a clearwell and mudwell, downflow filters may have a

higher structural cost than upflow continuous backwash

filters. However, since relatively small modules are used for

upflow continuous backwash filters, a greater number of

smaller filter cells normally is required to provide the same

surface area; this, too, can affect a structure’s cost. Whether

the influent and backwash piping and the valves associated

with downflow filters are installed outdoors or housed in

a building will affect the cost of a full installation. The 6 × 9

mesh media with a uniformity coefficient of approximately

1.35 and minimum sphericity of 0.8 tend to be more costly

than other media because only a limited number of sources

of media meet this specification. However, these media may

offer some operating cost savings by allowing more efficient

backwashing.

In addition to capital cost, it is important to consider

operating costs. The energy costs associated with back-

washing, air-scour, and nitrogen-release cycles must be

included, along with a proper accounting of the frequency

of these operations. The cost of “retreatment” of spent

backwash water also must be included: Filters using only 2%

of the forward flow for backwashing will have a lower cost

Figure 2. Block Underdrains

TETRA® T-Block™

8 W E & T

for treatment compared to

those that consume great-

er amounts of backwash

water. Finally, the ability

to optimize methanol dos-

ages can affect the operat-

ing cost significantly. Some

facilities have reduced their

chemical consumption as

much as 30% after imple-

menting more efficient con-

trol systems.

The significant differ-

ences in denitrification

experience, patent consid-

erations, and competitive

bidding laws can compli-

cate a municipal facility’s

efforts to procure a denitri-

fication filter. The features

associated with each type

of filter and each manufac-

turer should be evaluated

carefully before design and

procurement to identify the

advantages and disadvan-

tages of each. The differ-

ences between downflow

and upflow continuous-backwash filters are significant enough that the

type of filter should be selected before design. Procurement methods

that have been used recently for denitrification filter projects include

sole source, base bid, prequalification, and conventional bidding.

Christine deBarbadillo, P.E., is a process engineer in the

Charlotte, N.C., office of Black & Veatch Corp. (Overland Park, Kan.).

Robert Rectanus, P.E., is a senior project manager in the company’s

Gaithersburg, Md., office; Shannon Lambert, P.E., is a project

PAR

KS

ON

CO

RP.

In an upflow continuous-backwash filter, influent flows upward through the filter countercurrent to the movement of the sand bed.

About Severn Trent ServicesSevern Trent Services (www.severntrentservices.com), based in Fort Washington, Pa., is a leading supplier of water and wastewater treatment solutions. The company’s broad range of products and services is con-centrated around disinfection, instrumentation, and filtration technologies, pipeline analysis, rehabilitation and repair services, contract operating services and state-of-the-art residential metering products and services. Our international management services business provides support in all aspects of water and wastewater utility development and transformation. Severn Trent Services is a member of the Severn Trent Plc (London: SVT.L) group of companies. An international environmental services leader, Severn Trent is a FTSE 100 company.

For additional information about the TETRA® Denite® system email Ken Wineberg [email protected] or call 1.412.788.8300.

manager in the Nashville, Tenn., office; David

Parker, P.E., is a project manager in the Charlotte

office; Jeff Wells, P.E., is a project manager in the

Greenville, S.C. office; and Robert Willet, P.E., is a

project manager in the Raleigh, N.C., office.

Reprinted with permission from Water Environment & Technology, June 2005, by The Reprint Dept., 1-800-259-0470; #9856-0705