14th North American Waste to Energy Conference

May 1-3, 2006, Tampa, Florida USA

NAWTEC14-3189

An Assessment of the Capabilities of the Munters Zeolite Rotor Concentrator to Reduce VOC and Odor Emissions from a Municipal Waste Combustion Facility

James M. Secunde Maine Energy Recovery Company

3 Lincoln Street, P.O. Box 401 Biddeford, ME 04005 Phone: (207) 282-4127

Fax: (207) 282-8256 E-Mail: iim.secunde@casella.com

Abstract

Peter Krenitsky Munters Corporation - Zeol vision

79 Monroe Street, P.O. Box 600 Amesbury, MA 01913

Phone: (800) 843-5360

Fax: (978) 241-1220

E-Mail: oete krenitskv@munters.com

Maine Energy Recovery Company is a waste-to-energy facility, firing refuse-derived fuel ( R DF) in two B&W boilers to produce steam which is used to generate 2 2 M W of electricity. As part of its on-going effort to study odor generation and enhance their odor control system, Maine Energy discovered that a greater quantity of volatile organic compounds (VOC) are generated by the waste itself than had previously been estimated.

The VOCs that were found are primarily light alcohols, such as methanol, ethanol, and butanol, along with compounds such as acetone, methyl ethyl ketone ( M E K), benzene, toluene, xylene, and others. These compounds are generated from the operation of diesel-fueled equipment in the facility's tipping building, and from the decomposition of the waste itself. The VOC generation also has a strong seasonal component, where generation is highest in the warmer summer weather, and lowest in the depths of winter.

In the summer of 2005, Maine Energy undertook a pilot scale study of VOC control using a proprietary concentrator technology from Munters Corporation, Zeol Division of Amesbury, Massachusetts. A scaled-down version of their rotary zeolite concentrator was employed at Maine Energy over a six week period from July to September 2005. Numerous samples were taken at the inlet and outlet of the device, and several extended tests were conducted using Fourier-Transform Infrared ( FT I R) technology to search for specific organic compounds.

The results showed that the device reduced VOC, as well as odors, by approximately 85%, without the benefit of extensive fine-tuning of the device or the process during this limited run. The testing also revealed the need for extensive particulate removal at the inlet to the device, which would have a significant effect on cost efficiency.

Introduction

Maine Energy Recovery Company is a waste-to-energy facility located in Biddeford, Maine, approximately 30 minutes south of Portland, and 1.5 hours north of Boston, MA. Maine Energy receives 1100-1500 tons of municipal solid waste (MSW) per day, which it processes through several steps to produce refuse-derived fuel ( R DF). The R D F is used to fire two B& W boilers, which produce steam to drive a 22 MW

9 1

turbine/generator. The facility employs 85 people, and operates 365 days a year.

The facility was opened in 1987, and is located in the heart of downtown Biddeford, in a densely populated and developed area. While the immediate neighbors are mostly commercial and industrial properties, high-density residential areas are only a short walk away.

Copyright © 2006 by ASME

For many years, Maine Energy has been engaged in an intensive effort to identify the causes of perceived odors, and take steps to reduce and/or control them. In the course of several investigations over recent summer seasons, it was noted that there were unanticipated levels of volatile organic compounds (VOCs) present in the air being discharged from the facility. In 2005, Maine Energy undertook a project to study the applicability of a commercially available VOC concentrator technology, which is the subject of this paper.

Background

In 2000, the original operators of the Maine Energy facility, Kuhr Technologies, Inc ( KT I) merged with Casella Waste Systems of Rutland, Vermont. Upon assuming the operations of Maine Energy, Casella committed to an aggressive program to reduce the perceived odors in the community. One of the first issues identified was insufficient negative pressure in the tipping and processing areas of the facility. The existing system of drawing air from the tipping building for use in the boilers was not able to assure consistent negative pressure. In response, Maine Energy installed an extensive system of fans and ductwork to draw more air from these waste handling areas, handle it separately from the combustion air, and treat it prior to discharging it to the atmosphere. At the end of the air handling and treatment process were three cross-flow wet scrubbers, which use water to remove odorous compounds in the air.

In 2002, Maine Energy conducted a study using different chemical products in the scrubber water to determine their relative effectiveness in removing odor-causing compounds. As part of this testing, a number of bag samples of treated air were collected from the scrubbers and analyzed for a pre-determined list of organic compounds. In the course of reviewing this analytical data, it was determined that while the data did not allow for quantification of the VOC present. it was clear that the total mass of VOC was more than was anticipated when the scrubbers

Copyright © 2006 by ASME 92

were installed. Given that all previous design work, and subsequent testing, had been focused on odor removal, the existing data was inadequate for determining VOC concentrations and emissions.

In 2003, after consultation with the Maine Department of Environmental Protection, ( D E P) Maine Energy conducted a series of five testing events to more accurately measure VOCs in the air leaving the odor control scrubbers. These tests were conducted in February, April, May, August. and October of that year, in an effort to identify any seasonal trends. The tests were conducted as U S E PA Method 25A, and were run for continuous periods ranging from 48 to 60 hours.

Additional testing was conducted in the summer of 2004, at points upstream of the odor control scrubbers, and closer to the point of generation, which is the tipping building. These tests served to better quantify total VOC generation in the facility, and determine the removal efficiency of several additional odor control systems.

In 2005, Maine Energy contracted with Munters Corporation, Zeol Division, of Amesbury, MA to provide a pilot-scale version of their proprietary VOC concentrator technology for a short-term demonstration of its capabilities in the challenging environment of a solid waste processing facility.

Key Considerations

Through the summer of 2004, and in consultation with the Maine D E P, Maine Energy voluntarily developed a short list of VOC control technologies that seemed to merit investigation. Among them were several thermal processes (regenerative, and catalytic thermal oxidizers), solid treatment media (such as activated carbon), as well as the VOC concentrator. In the course of investigating these technologies, several issues proved to be significant considerations.

1. The air to be treated is a high-volume, low-concentration source.

In this instance, the volume of air to be treated is in the range of 20,000- 30,000 cubic feet per minute. The total concentration of VOC in the air stream is in the range of 100-300 ppm, of a broad range of organic compounds (see Table 1).

Table 1 - Typical voe species

methanol ethanol n-Butanol acetaldehyde methylene chloride methyl ethyl ketone 4-ethyl toluene l,l,l-trichloroethane acetone hexane

toluene benzene o,m,p xylene acetone 2-propanol tetrachloroethene heptane ethyl benzene tetrahydrofuran cyclohexane

Table 1 is not intended to be an exhaustive list of all compounds identified over three years of testing, but reflects the complicated mix of compounds that must be managed. Testing in 2005 showed that ethanol, butanol, and acetone make up roughly 80-90% of the organic compounds detected. With the exception of those three, the other Table 1 compounds are found at low part-per-billion levels.

2. VOC generation is higher in the summer.

As might be expected, New England is subject to the extremes of seasonal weather. Winters tend to be long and very cold, summers brief but hot, and the transitional seasons of spring and autumn are wet and highly variable. Through its experience, Maine Energy noted a significant rise in complaints of perceived odor in the hottest months of July and August, and a near total absence of complaints in the colder months from November through March (see Fig. 1).

While there are many possible explanations for this phenomenon, it was assumed that the primary cause was that the solid waste in the summer was received in a state of active decomposition and was generating

93

more odor-causing compounds, VOCs among them. Waste that was generated, stored, and transported at lower, or even freezing, temperatures is assumed to have limited biological activity.

Figure 1 - odor complaints by

month

120 100

80 60 40 20

0 >- c: III :::J :IE -,

3. Particulate the facility efficiency

� Cl Q. :::J -, c( CII en U >

0 0 z

-+-2001 __ 2002

2003 2004

loading and humidity in could impact control

It is not unexpected to find that solid waste processing facilities are dusty places. As a general statement, the particulate falls into three broad categories. The first is the fibrous material of the type that is released from dry paper as it is handled and processed. The second is the fine, dry dust that is released from sawdust, plastic grindings, and other similar sources in the waste stream. The third is the black soot that is generated from the diesel engines that power both the facility operating equipment, and the waste hauling vehicles that deliver the waste. These three types of particulate combine to form a dense and sticky material that quickly blinds over many types of particulate removal equipment. While the total PM in the tipping building is generally in the range of 4-6 mg/m3, the large volume of air that is to be handled makes particulate removal a significant concern.

Copyright © 2006 by ASME

Technology Selection

As a part of the on-going commitment to advance odor and VOC control efforts, Maine Energy reviewed the different technology types in light of the considerations identified above, with the additional consideration of cost effectiveness.

The various thermal technologies, while highly effective, have a high capital cost, and high energy costs. These devices also produce emissions of their own, which would require additional permitting. In our estimation, these technologies would not be appropriate for our situation.

Solid media, such as activated carbon, have a long history of use in the control of odors and the capture of organic compounds. However, this application would require very large amounts of media, in very deep beds, potentially 3 feet or more. This would entail considerable capital costs, and high material costs due to the quantities involved, and the frequency of material changes. Testing in 2004 on the l­inch beds currently in use as part of the odor control system demonstrated that while activated carbon reduces odors by 30-40%, total VOCs are reduced by less than 5%.

The concentrator technology seemed more suited for the testing that was contemplated. The technology is designed for situations where low concentrations of VOCs are present in large air volumes, and has energy requirements far lower than those necessary for thermal technologies. Full-scale versions of the Munters concentrator system are in use world-wide, including two large applications at semiconductor plants in the greater Portland, Maine area. The technology is well established, and has a history of excellent performance. While most of the current applications are in cleaner environments than are typically found in waste handling facilities, the application for Maine Energy appeared plausible. The headquarters and manufacturing center for the Munters Zeol Division is less than 2 hours' drive from Biddeford, and the company was motivated to develop and deliver a pilot-

Copyright © 2006 by ASME 94

scale version of their device for this project. In January of 2005, the development of the pilot scale unit was initiated.

The Munters VOC Concentrator

Figure 2 - Drawing of Munters Pilot-Scale

VOC Concentrator Unit

The Munters Zeol® Rotor Concentrator is a pollution control device designed to remove volatile organic compounds (VOC) from air streams. The system typically includes two process steps:

1 ) Concentration of the VOC using a Hydrophobic Zeolite Rotor, and

2) Post treatment of the concentrated VOC

As solvent laden air is drawn through the HoneyCombe rotor, VOCs are removed from the air by adsorption onto the hydrophobic zeolite. The cleaned air passes through the rotor and is discharged to the atmosphere.

The Zeol Rotor turns at a speed of one to six revolutions per hour, continuously transporting adsorbed VOCs into a desorption sector, and returning regenerated zeolite to the process air stream. In the desorption sector, the adsorbed VOC are removed from the zeolite with a small stream of heated air. The volume of desorption air containing the

concentrate is generally less than 10% of the incoming volume.

As it exits the desorption sector, the rotor is cooled with a small portion of process air. This air stream, referred to as the cooling air, is captured in an isolated plenum, heated, and returned to the rotor's desorption sector to remove the VOC.

Post-treatment of the concentrate typically consists of a thermal oxidizer, which is integrated with the concnetrator. As an alternative to the use of an oxidizer or other thermal device, the concentrate at Maine Energy could be ducted to the boilers for use as combustion air, while also destroying the VOC content of the concentrate.

Testing Program

Given that the effectiveness of the concentrator technology is well-established in other industries, the major goal of this testing program was to assess its effectiveness in a waste handling operation, and identify any limiting factors that may impact its applicability in such a facility.

In the course of planning for this test, Maine Energy considered a number of different scenarios. In the final analysis, it was decided that the most appropriate place to perform the testing was in an area nearest the tipping building. Some of the factors that were considered:

1. Testing in the area that appeared to generate the most VOCs and odors

2. Locating the device in an area that would be accessible for both set-up and sampling

3. Minimizing distances to be run with ductwork and wiring

4. Providing an environment that would present the most challenging dust and humidity

5. Protecting the device from the weather, and from operating equipment

95

Since the unit was intended to be utilized only through the hotter summer months, a great deal of information needed to be gathered in a relatively short period of time. Initially, a number of categories of data were identified that would be useful in determining the effectiveness and utility of the system:

1. Ambient temperature, humidity and inlet air flow velocity

2. Total hydrocarbon inlet and outlet concentrations

3. Analytical testing for specific organic compounds

4. Odor sampling at inlet and outlet

5. Particulate loading at the inlet through filter weight determination

Ambient temperature, humidity and velocity were to be measured at least once each weekday using a "hot-wire" multi-function anemometer. These readings would be recorded manually, and logged in a spreadsheet. A hand-held photo-ionization detector ( PI D), configured to display results expressed as propane, was used to measure total hydrocarbon concentrations at the inlet and outlet of the concentrator, with several readings also taken of the "concentrate". In addition to this analysis, two 48-hour continuous tests were scheduled to be conducted to track performance of the unit over an extended time. A PCC, Ltd. of Greene, M E was contracted to perform this testing. A PCC proposed to use a flame-ionization detector ( FlO) to determine total hydrocarbons, as propane. During the second of the two events, A PCC would utilize a different analytical device, a Fourier-Transform Infra­red analyzer ( FTI R) to analyze the inlet and outlet air for concentrations of specific organic compounds at levels above 1 part­per-million (ppm). For compounds that may be present at levels less than 1 ppm, samples would be collected utilizing specially-prepared evacuated metal canisters, that were analyzed by a separate laboratory using US E PA I S method TO-15. This method quantifies a standard list of 60

Copyright © 2006 by ASME

organic species, and at Maine Energy's request, reported the top 20 tentatively identified compounds (T ICs).

Odor samples from the inlet and outlet of the concentrator were collected on eight separate occasions. The inlet/outlet samples were collected simultaneously in 1 2-liter Tediar™ sample bags that had been purged prior to sample collection. These samples were then sent to be evaluated by an odor panel conducted by Odor Science & Engineering, of Bloomfield, CT.

The pilot-scale concentrator unit was delivered to the site in mid-July 2005, and after approximately a week of set-up and start-up testing, was placed in service on July 2 2. The first continuous VOC tests took place on August 10-12, and the second run was conducted September 6-8. Sampling with evacuated canisters took place during these test events. Daily monitoring was conducted for temperature, humidity, air flow, and VOC concentrations at the inlet and outlet of the concentrator. The two­stage, pleated inlet air filter arrangement was inspected frequently, and filter changes were noted. Due to the size and configuration of the filters, however, as well as variations in air flow, data on particulate accumulation could not be obtained.

Following the completion of the VOC testing on September 8, the unit was taken off-line, disconnected from the ductwork and electrical service, and returned to Munters.

Testing Results

1. Odor

The results from the odor panel analysis are given in units of dilutions-to-threshold (d/t), which is the number of dilutions necessary to reduce the odor in the sample to the point where it is no longer detectable by the panelists. The results are shown below in Table 2.

Copyright © 2006 by ASME 96

T bl 2 a e - Odor sampling results expressed as d/t

Sample 10 Inlet Outlet % Reduction

"A" (B/11 am) 797 149 81.3 " B" (B/11 pm) 1092 177 83.8 "C" (B/1B am) 2 314 177 9 2.4 "0" (B/1B am) 3 276 298 90.9 "E" (9/1 am) 638 177 72.3 " F" (9/1 pm) 988 137 86.1 " G" (9/7 am) 867 162 81.3 "H" (9/7/pml 797 193 75.8

As demonstrated by the results in Table 2, the VOC concentrator was very effective in reducing odor levels in the tipping building air. However, it is very difficult to directly correlate these results with data from the VOC data gathering efforts, for several reasons:

1 . Odor samples were collected over very short periods of time, generally less than 10 minutes.

2. The number of odor samples is limited, as compared to the 48 hours of data from the VOC sampling effort

3. Measured VOC concentrations varied widely, even on a minute-to-minute basis

While the focus of this trial was on VOC reduction performance, odor control was of Significant interest to Maine Energy. Personal observations made during the testing clearly indicated that the outlet air has significantly lower odors, while the odors in the "concentrate" were quite potent. The results shown in Table 2 confirmed these more casual assessments.

2. Volatile Organic Compounds

The results from the continuous F I D monitoring conducted by A PCC are summarized in Tables 3 and 4, which represent the two 48-hour testing events. The data is expressed as total non-methane hydrocarbons (N MHC), as propane. Graphical representations of the data are in Fig. 3 and 4. The VOC concentrations over the 48-hour periods varied widely and frequently. Subsequent attempts to correlate the various peaks to specific activities or conditions in the tipping building were

unsuccessful. The data clearly indicate the need to conduct such testing over

extended periods, and not to rely on short­term, "snapshot" data.

In the graphics contained in Fig. 3 and Fig.

4, the sheer quantity of data involved

requires a certain amount of editing

flexibility. The uppermost line reflects

removal efficiency on a mass basis. Inlet

concentrations are represented by the line

in the middle of the graph, and outlet

concentrations are indicated by the lower

line. Gaps in the data represent either

pauses for instrument calibration, or periods

when sampling was suspended for

concentrator unit downtime.

Table 3 - Summary of VOC removal efficiency, August 10-12

Inlet ppm Outlet ppm Flow In Flow Out VOCln VOC Out Removal (scfml (scfm) Clb/hrl (lb/hr) %

Average 159 28 654 628 0.236 0.039 83.5% Maximum 301 54 741 716 0.455 0.078 89.7% Minimum 30 5 591 542 0.049 0.008 74.6%

Figure 3 - Total hydrocarbon concent rations and rem o val efficiency, August 10-12

1 3S0 r-----------------------------;:::;:;;:::======::::::1 00%

2S0t-----------------�_.���-------------� 70%

ppmv 200t----ir----���_fTH_r---�--�l+-----�ht-----------�-� 60%

SO%

40%

100 t----t--V--------------------------------+---JIV--L..!...----f 30%

20%

10% S0t-�r.ftjN�----------_A-------4.�-.��--------------------�

---1 '\(' I'M. J'yvl �� 8/10/0': S-:-:12:-::0":'0 ..:;...-.;=...:.....-<--8�/ 1 -1I-0S -0 -:00---�- -8� /1�1/0- S�1 2 ..... : 0-0---- �8-/1>-2/� OS

-0-:0 -0 -�---8-/1 -2/-0S�1-2 : -00-J

0%

97 Copyright © 2006 by ASME

Table 4 - Summary of VOC removal efficiency, September 6-8

Inlet ppm Outlet ppm Flow in Flow out Mass in Mass out Removal (scfm) (scfm) (Ib/hr) (Ib/hr) %

Average 152 23 696 679 0.242 0.036 85.5% Maximum 286 51 777 731 0.462 0.081 92.3% Minimum 45 5 609 611 0.068 0.008 75.4%

Figure 4 - Total hydrocarbon concentrations and removal efficiency, September 6-8

350r;::====:::;-

----------------------------, 100.0%

- Inlelppm -Oullelppm

-Efllciency('Kowfw) � � 300 �����-A-::----=-----=---*����\rJ-=-----l

90.0%

80.0%

250 �------II---------I--'----------------------I--� 70.0%

60.0% 200�----����r----�_+����--��� 11__------_I'1__�

ppm v

50.0%

150+----.-i------�----1�---L�-�-11R_1_�-1_�-_+-�.-1 40.0%

100 �-.I_-__t_----__t__tl \--If-------i---f-\ ,----¥--t----I-I 1.1-------1 30.0%

20.0%

50 +--------------1�-�--�----��------;,-�

A\fJVj AA�� 10.0%

.j...-;----+----....... �...-.-.......--......... ----�-----___ _...-_-� ____ ...... O.O% 9/6/05 12:00 9/7/05 0:00 9n10512:00

Overall, the continuous VOC data leads to several interesting observations:

1. There are no obvious correlations between facility operating levels and VOC generation. For example, the graphical data in Fig. 4 show a distinct, extended decrease in the inlet VOC concentrations during the night of September 6-7. This is what would be expected, since the tipping building is much less active in the evening hours when there is no incoming waste. However, this trend does not occur as obviously on the succeeding night, nor does it appear in the August data. This may

Copyright © 2006 by ASME 98

9/8/05 0:00 9/8/05 12:00

reflect the need for closer monitoring of operational activities to detect a cause for these fluctuations.

2. Observation of the graphical data in Fig. 3 and Fig. 4 shows that concentrator removal efficiency reacts inversely to relative inlet VOC concentration. Peaks in VOC inlet concentration are consistently matched by decreases in removal efficiency. Several potential causes for this have been suggested:

• The efficiency of the concentrator unit is, in part, determined by the rotational

speed of the zeolite rotor and the quantity of air being treated. G iven that the speed and air flow of the unit were f ixed during the course of this test, the zeolite may have been br iefly overwhelmed.

• The peaks in total VOC may be related to a short-term increase in compounds that are not readily removed by the concentrator, such as methanol, or acetaldehyde.

Data from the FT I R analysis focused on what were known to be the predominant compounds in the a ir stream, specifically, methanol, ethanol, n-butanol, and acetaldehyde. Taken together, these compounds often account for as much as 80-90% of the total organ ic compounds found in the t ipp ing building. Table 5 l ists the average inlet and outlet concentrations of these compounds on the days when sampling took place.

Table 5 - Concentrator control efficiencies for selected organic compounds

methanol Ethanol n-butanol acetaldehvde T HC Sept. 6

Inlet ppmv 4.4 15 Outlet ppmv 3.9 0.8

Sept. 7 InletQRmv 5.7 17.5

Outlet ppmv 4.9 1.5 Sept. 8

Inlet ppmv 4.9 19 Outlet ppmv 4.5 2.9

Calculated control efficiency by concentration 11.9% 90.2%

Interest ingly, the data shows a strong preference in the zeolite for certain organic species. In this case, ethanol and n-butanol are removed with relatively h igh efficiency, while methanol and acetaldehyde are controlled at much lower levels. This data also shows a total hydrocarbon reduction that is consistent w ith data from the F I D.

Additional analyses were conducted on the canister samples that were collected concurrently w ith the F I D/ F T I R efforts. Samples were collected from the inlet and outlet of the concentrator over three-hour per iods, and analyzed using U S E PA's Method TO-15. These samples were targeted to identify compounds at concentrations less than 1 part-per-mill ion. The removal efficiency data for these organic species was typically above 80%, as shown in Table 6.

17.3 0.1 163 4.5 0.0 26.5

17 1.5 182 3.7 1.0 26.1

17.2 2.1 159 4.6 1.6 3 2.8

75.1% 27.5% 8 2.9%

T bl 6 C tr I Effi . a e - on 0 IClenc) : M th d TO 15 e 0 -

Compound Avg. Avg. % Inlet Outlet Removal (ppb) (ppb)

styrene <68 <15 77.0% toluene 883 121 86.3% Ethyl benzene 116 16 85.9% M.p-xylene 377 49 87.0% hexane 106 16 85.4% heptane 99 11 89.3% 2-propanol 967 150 84.3% 2-butanone 957 150 84.3% tetrahydrofuran 136 <5 96.1% acetone 2167 300 86.2%

As discussed previously, monitoring of the process was also conducted on a regular basis through the use of two hand-held instruments. A portable photoionization detector ( P I D) was used to measure inlet and outlet VOC concentrations, expressed as propane. A "hot­wire" anemometer was also used to gather

99 Copyright © 2006 by ASME

data on air flow, velocity, relative humidity, and temperature of the inlet air. While the data from the anemometer was reliable and consistent throughout the test period, the P ID yielded data late in the test that appeared to be an under-reporting of the VOC concentration. The data generated late in the test program appeared to be roughly a factor of ten less than what had been reported earlier in the program. The most likely cause was a deterioration of the lamp in the device. The data was still collected, and calculations of control efficiency showed good correlation with previous data. The data is summarized in Table 7. The data on VOC reduction was consistent with that collected by both the F ID and FT I R, giving a high level of assurance that all three devices were giving good data, and that the overall result of 80-85% control efficiency is accurate.

It is worth noting a significant deviation from the monitoring plan as it was envisioned at the beginning of the test. In the original scenario, the two F ID/ F T I R test events were designed to represent an "as-delivered" and an "as­optimized" condition of the concentrator unit. The August test event was conducted with only minimal set-up and adjustment by Munters, and documented a baseline condition. In the interim period prior to the September test, the intention was to have Munters conduct additional trials with the concentrator, so that its operating characteristics could be optimized for the particular conditions found at Maine Energy. However, due to a miscommunication between Maine Energy, Munters, and the testing contractor, this period of optimization did not occur. As such, there remains the possibility that the removal efficiency of the concentrator could have been improved, and that this data represents a minimum efficiency in this application.

a e -T bl 7 M I b anua 0 f serva Ions 0 fVOC concen a or opera Ion tr t f Date Time Inlet Inlet Inlet T Inlet R H VOC in VOC %

velocity flow (deg. F) (%) (ppm) out reduction (fpm) (cfm) (ppm)

July 29 0930 740 403 87 64 190 18 90.53 Aug 1 0940 690 376 82.5 69 210 3 98.57 Aug 2 1000 765 417 86.5 65 Aug 3 1030 725 395 90.1 59 140 20 85.71 Aug 4 0910 705 384 81.5 61 112 22 80.36 Aug 8 1420 1170 637 95.1 57 Aug 9 1030 1110 605 91.2 63 Aug 10 1000 1120 610 91.5 54 Aug 17 1000 1120 610 86.7 69 Aug 18 1010 1090 594 83.2 50 110 11.5 89.55 Aug 18 1345 1080 588 85.9 47 145 9.5 93.45 Aug 19 0945 980 534 82.2 65 215 25 88.37 Aug 22 0900 1090 594 84.2 60 Aug 23 1300 1050 572 88.5 45 23.3* 2.1 * 90.99* Aug 24 1345 1060 577 86.2 5 2 33.4* 3.8* 88.62* Aug 26 1415 1090 594 90.4 51 Aug 29 1330 1100 599 88.5 68 Aug 30 1030 1060 577 88.7 71 4* 1 * 75* Sep 1 1530 1010 550 91.5 62 25.3* 2.3* 90.91* Sep 2 0900 900 490 85.2 69 26.5* 2.7* 89.81* Data marked with an asterisk (*) IS considered suspect due to equipment deterioration

Conclusions

As a general statement, the goal of the testing program was not to achieve a certain level of removal efficiency, or to prove or disprove the

Copyright © 2006 by ASME

effectiveness of the technology. The effectiveness of the Munters concentrator in removing low-concentration VOCs from large air volumes is well-established. The objectives

100

of this testing were to determine if the operating conditions found in a waste handling facility would interfere with the operation of the device, or negatively impact the device over time, and to identify any factors that would limit the applicability of this technology to such a facility.

Based on the similarity of the VOC removal efficiency results between the August and September test events, and the relative consistency of odor removal results, one can conclude that the rotary zeolite concentrator was able to maintain its effectiveness over the duration of this six-week test program. As noted previously, with additional effort to fine-tune the operating parameters of the concentrator, it is likely that a higher level of removal effectiveness could be achieved and maintained.

Potential limiting factors were evaluated during the test program, and many were found to be of no significance. The variations and levels of humidity and ambient temperature had no apparent effect on operational reliability or removal efficiency. While particulate in the inlet air was controlled by the two-stage pleated filter arrangement in the pilot scale unit, the level of effort necessary to maintain the filters indicate that particulate control would be a major design consideration for a full-scale installation. The need for filter replacement was monitored by the differential pressure across the filter arrangement. When the air flow rate through the unit was increased in early August the increase in pressure differential, and thus the frequency of filter replacement, increased significantly. Unfortunately, particulate loading as mass or concentration could not be determined in this testing program, but clearly is sufficient to warrant considerable engineering and planning effort in a full-scale installation.

The remaining consideration is that of cost­effectiveness. Given the condition of a low­concentration target compound in a large volume of air, the cost per ton of pollutant removed by any technology would be expected to be high. Preliminary calculations performed by Maine Energy indicated that the cost of a full-scale concentrator installation would be approximately $14,000 per ton of pollutant removed, without consideration of any of the necessary design, procurement and installation costs of a particulate removal

101

device. This presumes that the concentrate could be directed to the boiler for post­treatment as combustion air. This compares to a cost of roughly $19,000 per ton for a thermal oxidizer, again without including the costs of particulate removal. Either cost is well above what would generally be considered to be cost­effective.

Costs notwithstanding, the VOC concentrator proved to be an effective technology for controlling VOCs and odors in a waste handling environment.

The Authors

Jim Secunde is Environmental Manager at Maine Energy Recovery Company in Biddeford, Maine, having worked at the facility since 200 l. Mr. Secunde received his Bachelor's Degree in Environmental Engineering Technology in 1986 from the University of Dayton in Dayton, Ohio, and has been employed in the field of environmental compliance for nearly 20 years. In his work at Maine Energy, he has developed extensive experience in odor generation and control, in addition to his expertise in air and water pollution control developed in his work as a regulator, consultant and project manager.

Pete Krenitsky is the Sales Manager for Munters

Zeol, where he has worked for seven years. Mr.

Krenitsky is responsible for applications development and manages a small sales force

that delivers product worldwide. Previously, he

worked for TRC Environmental, a leader in

emissions measurement for the waste-to-energy

industry.

References

Powell, John H., " Munters VOC Concentrator Efficiency Test, Maine Energy Recovery Co., Biddeford, ME, August-September 2005". A PCC Ltd., Greene, ME

Grumley, Gary K.,and O' Brien, Martha, "Odor Panel Analysis, August & September 2005", Odor Science & Engineering, Bloomfield, CT.

Copyright © 2006 by ASME