feature - pubs.awma.org

14 em january 2007 awma.org

emfeature

Locomotives are important sources of nitrogen oxides (NOx), fine particulate matter (PM2.5), and diesel particulate matter (DPM) emissions. The U.S. Environmental Protection Agency (EPA) has declared DPM a probable toxic air contaminant and, in 1998, California established DPM as a toxic air contaminant with potency factors for determining human cancer risk. Wherever major locomotive-related activities occur, the potential exists for elevated cancer risk, especially in areas adjacent to popula-tions. One such setting is in the community of Roseville, CA, approximately 15 miles east of Sacramento, where the Union Pacific Railroad operates the J.R. Davis Rail Yard (Yard). According to data compiled by the California Air Resources Board (CARB) for the period of December 1999 through November 2000, approximately 31,000 locomotives stopped at the Yard, while another 15,000 locomotives passed through the facility without stopping.1 During this period, CARB estimated that DPM emissions from locomotive activities were between 22 and 25 tons per year (t/yr).

The Yard, which has been in existence since 1905, cov-ers almost 1000 acres, and spans the border of two counties: Sacramento and Placer. Downtown Roseville is immediately adjacent to the facility to the southeast, and there are resi-dential areas adjacent to the north and south of the facility. The Yard performs several functions. It is a classification

yard, which means that cars from incoming trains are dis-connected and then reconnected to other trains bound for specific locations across the country. These operations occur in the “hump and trim” section of the Yard. In this process, a locomotive pushes an incoming train of connected cars, in reverse, up a slight incline. The cars are individually discon-nected at the apex of the incline, where each separated car then moves downhill by gravity. A series of parallel tracks with switches allow the operator to direct the moving car to the designated track for reassembly into another train. The facility also serves as a refueling, maintenance, service, and

repair facility for locomotives. In some of the maintenance and repair functions, locomotives are tested under load, and so emissions from diesel under both load and nonload condi-tions occur. A layout of the facility, its key operations, and the community surroundings are shown in Figure 1.

The expansion of activities at the Yard in the late 1990s led to community concerns about the potential air toxic risk from DPM emissions. The Placer County Air Pollution Control District (PCAPCD), which oversees the Roseville area, responded by requesting that CARB conduct a risk assessment of the Yard’s DPM emissions. A multi-year study was completed in October 2004.1 In conducting the study, CARB studied specific activity data during a 12-month period,

The challenge will be to reduce DPM emissions by at least 25% , while the volume of traffic at the Yard is increasing.

Melvin D. Zeldin is an environmental consultant in Sparks, NV. Yushuo Chang is the planning/monitoring

manager and Thomas J. Christofk is the air pollution control officer with Placer County Air Pollution Control District, Auburn, CA. David E. Campbell is an assistant research

scientist and Eric M. Fujita is a research professor with the Desert Research Institute, Reno, NV. Richard J. Countess

is a senior partner with Countess Environmental, Westlake Village, CA. E-mail: [email protected].

Copyright 2007 Air & Waste Management Association

awma.org january 2007 em 15

December 1999 through November 2000. From these data, CARB calcu-lated DPM emissions from facility operations. Spe-cifically, the results showed that approximately 50% of emissions were derived from moving locomo-tives, 45% from idling locomotives, and 5% from locomotive testing; hourly emissions rates of DPM were uniform throughout a 24-hr period, ranging between 5 and 6 lb/hr; total facility DPM emis-sions were estimated to be between 22 and 25 t/yr; and approximately 75% of the DPM emissions in the servicing area were derived from idling locomotives.

Using risk assessment modeling based on 70 years of exposure, elevat-ed cancer risk levels of 100–500 in 1 million were found to occur across an area of 700–1600 acres, affecting a population of 14,000–26,000; and elevated risk levels of 10–100 in 1 million were found to occur over 46,000–56,000 acres, affecting a population of 140,000–150,000. Data sets from the CARB Roseville Air Monitoring Station (1.5 km from the facility) and McClel-lan Air Force Base (10 km from the facility) were used. The modeling results are shown in Figure 2. The areas with the highest cancer risk levels are located near the service/repair and hump/trim operations.

Cooperative agreementWithin two months of the release of the CARB risk assess-ment, PCAPCD and Union Pacific Railroad entered into a groundbreaking agreement to reduce DPM emissions at the Yard. Under the agreement, Union Pacific committed to a mitigation plan designed to achieve a 25% emissions reduction from the 1999–2000 baseline period1; a grant program, providing a minimum of $150,000 toward local projects designed to reduce DPM from other activities in the Roseville area; and an air quality monitoring program around the Yard to be conducted over a three-year period (2005–2007), with assistance in the amount of at least $100,000.

mitigation planSince 2000, there has been an increase in rail-related traffic through the facility. The challenge will be to reduce DPM emissions by at least 25% from the baseline level, while the volume of traffic at the Yard is increasing. The 25% goal needs to be attained by the end of 2007, and methods to

calculate the reductions are being developed at this time. The ambient monitoring data obtained from the Roseville Rail Yard Air Monitoring Project (discussed in more detail below) will be used to track the trend in emissions reductions over the three-year period of the agreement.

The mitigation plan has four areas of focus: (1) the reduction of unnecessary idling; (2) the introduction of low-sulfur diesel fuel; (3) switcher locomotive fleet replace-ments/upgrades; and (4) the installation of control equip-ment. The first two areas (idling and fuel) address emissions reductions throughout the entire facility, while the latter two areas (switcher fleet upgrades and the installation of control equipment) target the highest sources of emissions, namely those areas adjacent to the service/repair and hump/trim operations. Each of these four focus areas is described in more detail below.

Reduction of Unnecessary Idling — CARB found that idling locomotives accounted for 45% of the total DPM emissions at the facility.1 The reduction of unnecessary idling should yield both emissions reductions and fuel savings. Union Pacific began implementing operational policy changes in 2001 to reduce idling emissions.

Low-Sulfur Diesel Fuel — In November 2004, CARB passed a regulation that requires locomotives that operate at least 90% of their time in California (intrastate) to use low-sulfur diesel fuel (i.e., maximum 15 parts per million [ppm] sulfur content and a 10% aromatic limit) by January 1, 2007. These intrastate locomotives currently consume 15% of the total locomotive fuel dispensed in California. The use of 15-ppm sulfur diesel

Figure 1. Aerial view of the J.R. Davis Rail Yard and location of key operations.



fuel is estimated to reduce PM emissions by 16% reduction and NOx emissions by 6%.2

Under current federal law, railroads are permitted to use nonroad diesel fuel with a sulfur limit of 5000 ppm in their engines. The use of low-sulfur diesel fuel for locomotives and marine vessels will not be phased in until the end of 2012. According to this schedule, Union Pacific’s interstate line haul fleet would not be mandated to use low-sulfur fuel for another six years. However, a recent memorandum of understanding has addressed this situation. Under the agreement, the two nationwide railroad companies operating in California—Union Pacific and Burlington Northern Santa Fe—have agreed to ensure that by January 1, 2007, a minimum of 80% of the diesel fuel supplied to locomotives fueled in California meet the state’s low-sulfur standard.

Switcher Fleet Replacement/Upgrades — The switcher fleet of locomotives operates throughout the Yard, but most of the emissions occur during the hump and trim operations. These switcher locomotives are typically older, lower horsepower models. Four locomotives are assigned to the hump operations, two handle trim duties, and two

additional locomotives serve as backup units. PCAPCD and the Sac-ramento Metropolitan Air Quality Management District (SMAQMD) have developed a project, in concert with Union Pacific, to replace one of the switcher locomotives used in the Yard’s hump and trim operations with a new Gen-Set Tier 2 model. It is estimated that the annual emis-sions reductions from replacing the old locomotive will be 25.16 tons of NOx, 0.45 tons of PM10, and 1.32 tons of hydrocarbons emissions. The new Gen-Set switcher locomotive should be operational in 2007. PCAPCD and SMAQMD are planning additional replacement units commensurate with future funding, until the entire fleet is replaced.

Installation of Control Equipment — The maintenance functions (i.e., diagnosis, service, repair, and testing) occur in an area of the facility where locomotives are generally stationary for long periods with their engines idling and occasionally running under load. This activity is a major source of DPM emissions that needs to be controlled. A team composed of members from PCAPCD, SMAQMD, Union Pacific, and Advanced Clean-up Technologies (ACTI) developed a proof-of-concept demonstration project to control emissions, which

involves the capture of locomotive exhaust with a collection hood or bonnet and rerouting it into air pollution control equipment. The project was awarded funding in August 2005 through EPA’s West Coast Diesel Collaborative, a pub-lic–private initiative to reduce diesel emissions. Recently, both CARB and SCAQMD were added to the demonstra-tion project team. Each party is contributing either dollars or “in-kind” services, or both. The total cost of the project is estimated at $1.4 million.

The project is structured in two phases. The first phase involved the locomotive interface design, test location definition and design, development of the test protocol, and acquisition of the locomotive interface hardware (i.e., the Advanced Locomotive Emission Control System [ALECS] from ACTI). The second phase involves erecting the ALECS at the test site, testing two different types of locomotive using the test protocol, disassembling the ALECS and shipping it back to ACTI, and preparing a final report. Phase I began in September 2005, and was completed in early 2006. Phase II started in June 2006, and testing began in September 2006. The ALECS equipment installed at the Yard is shown in Figure 3.

Figure 2. Estimated cancer risk from the rail yard (100 and 500 in 1 million risk isopleths). Notes: Solid line = Roseville Met Data; dashed contour lines = McClellan Met Data; urban dispersion coefficient, 80th percentile breathing rate; all locomotives’ activities (23 t/yr); modeling domain = 6 km x 8 km; resolution = 50 m x 50 m.



grant programUnion Pacific has agreed to provide grants totaling $150,000 to achieve a 1-ton DPM emissions re-duction from non-rail yard sources in the Roseville area, such as heavy-duty on-road and off-road equipment. In 2005, PCAPCD used $50,000 in grants from Union Pacific toward the district’s Clean Air Grant program to retrofit four Roseville City refuse trucks with emis-sions control systems. It is estimated that this retrofit will reduce DPM and NOx emissions by 0.2 and 2.73 tons, respec-tively, over the seven-year service life of these vehicles. In 2006, the PCAPCD used $100,000 in grants from Union Pacific to replace an old school bus with a new low-emissions school bus for the Roseville Joint Union High School District. Assuming the new bus has a 20-year service life, it is estimated that DPM emissions will be reduced by 0.6 tons and NOx emissions by 9.4 tons over the life of the bus. Between these two projects, DPM emissions will be reduced by approximately 0.8 tons.

roseville rail Yard air monitoring projectThe third component of the cooperative agreement was an air monitoring study. A technical advisory committee (TAC) was established to provide technical review and oversight for the project. Members of the TAC included representa-tives from the sponsoring agencies, as well as the California Office of Environmental Health Hazard Assessment, the University of California at Davis, and the Desert Research Institute (DRI). The TAC was instrumental in providing critical guidance to the design and implementation of the field study. Additional funds for equipment and field support were contributed by SMAQMD and EPA Region IX, with laboratory support from SMAQMD and field audit support from CARB.

The key objectives of this study were to determine local-ized air pollutant impacts from the Yard; verify, if possible, the effectiveness of mitigation measures being implemented by Union Pacific over the three-year period; provide ambient measurement feedback to the local community; and develop a more detailed and specific database for further modeling analysis.

Monitoring and Analysis. One of the difficulties with filter-based sampling is that it is an integrated sample over a period of time with varying wind directions. The project used continuous black carbon (BC)-measuring aethalometers, continuous PM2.5 beta attenuation monitors (BAMs), and co-located filter sampling to compare 24-hr average hourly results to Federal Reference Method (FRM) samplers. In addition, continuous NOx analyzers helped with diesel plume detection, as simultaneous peaks in both NOx and BC would, in this instance, strongly indicate the presence of a diesel emissions plume.

Critical to the design of the monitoring program was

Figure 3. ALECS equipment installed at Roseville Rail Yard.

Figure 4. Location (•) of upwind and downwind monitoring sites.



the location of upwind and downwind monitors. While the CARB risk assessment indicated areas of greatest risk to cancer, the TAC recommended that a pilot field sampling program was needed to confirm that elevated levels of am-bient PM2.5, BC, and NOx could be detected at meaningful levels. PCAPCD contracted with DRI to conduct a pilot study using an instrumented mobile van. Measurements included BC by photo acoustic analyzer, total NOx, carbon monoxide, carbon dioxide, volatile organic compounds by photo-ionization detection, particulates by sampler, and other ancillary data. The mobile van operated primarily during the period 10:00 p.m.–6:00 a.m. to coincide with the optimum summertime southeasterly prevailing winds (which blow perpendicular to the Yard orientation). The findings showed substantial concentrations of BC and NOx near the downwind Yard service area, and very low concen-trations on the upwind side.3 Using these results, two pairs of upwind and downwind monitoring sites were installed (see Figure 4).

Meteorological data for the area showed the greatest nighttime wind persistence during the summer months, when influences from synoptic scale weather events are at a minimum. For the first year of monitoring (2005), a period of mid-June to mid-September was selected as the primary sampling period; however, due to logistical startup difficulties, the second upwind-downwind pair of sites did not become operational until August 2005. As a result, the

sampling period was extended through mid-October.Quality assurance was conducted according to proce-

dures specified in a Quality Assurance Project Plan that was submitted to and approved by EPA. This included co-located monitoring before and after the summer sampling period and a field audit conducted by CARB. Data valida-tion included several levels of review. Of note are the data collected by the aethalometers. There are several operational features of the aethalometer that can affect comparability of data from multiple instruments. Baseline measurements are made after each tape advance, resulting in a 15-min gap in the data. The advance occurred when instruments reached a certain filter loading, as set by the manufacturer, so that downwind instruments cycled more frequently than upwind sites. The aethalometer has been shown to over-predict BC concentrations on a fresh filter and under-predict BC concentrations on a loaded filter.4 Aethalometer data are also known to be strongly affected by electronic noise spikes that create exaggerated increases or decreases in individual measurements of light attenuation. Extensive analyses by DRI showed that the measurement uncertainty created by these problems could be reduced to acceptable levels when data are aggregated for at least six hours.4

DRI conducted a complete analysis for the first year of data.5 To determine the conditions upon which up-wind–downwind analyses are appropriate, three sets of screening criteria were established: winds need to be from

Figure 5. Comparative results for upwind (P and V) and downwind (D and C) monitors for (a) BC, (b) PM2.5, (c) NO, and (d) NOx. P = Pool; V = Vernon; D= Denio; C = Church.



a semi-circular arc between 45 degrees (i.e., northeasterly) through 225 degrees (i.e., southwesterly); only winds of 1–5 mph were used to eliminate calm or windy conditions; and only overnight hours from 10:00 p.m. to 5:00 a.m. PST were used. This is the time frame when the winds blow most consistently across the rail facility directly from the upwind to the downwind locations, and therefore, emissions from the rail facility can most readily be detected.

Results. DRI evaluated the differences between upwind and downwind site concentrations of BC, PM2.5, NO, and NOx. The results are shown in Figure 5. Figure 5a shows that the mean concentrations for BC at both downwind sites (Denio and Church) are significantly higher than at their corresponding upwind sites (Pool and Vernon). Uncer-tainty estimates are also indicated. Also shown are the dif-ferences between the upwind and downwind pairs to show the presumed impact from the rail yard facilities. For BC, these are over 1.5 µg/m3. From a statistical standpoint, the differences in mean concentrations between upwind and downwind sites are significant above the 99% confidence level. Also shown in the rightmost bars are the comparisons of the upwind and downwind sites, which indicate both pairs of upwind and downwind sites consistently reflect the same conditions.

Figure 5b shows the results for PM2.5 mass. While the downwind sites have levels that are statistically higher than the upwind sites, these differences are not as pronounced as for BC. This is because PM2.5 mass is a regional pollutant that affects both upwind and downwind sites. Nevertheless, the differences between upwind and downwind sites are approximately 7–12 µg/m3.

Figure 5c shows the results for NO. NO is a good indi-cator of fresh NOx emissions. This chart may be the most indicative that the downwind sites are picking up the emis-sions from the rail yard facility. While downwind sites show NO concentrations of approximately 130 ppb, the upwind sites register concentrations of less than 10 ppb.

Figure 5d shows the results for NOx. While the results of Figures 5c and 5d are very similar, there are some interest-ing differences. The downwind sites show a very high per-centage of NOx as NO, meaning these sites are dominated by fresh emissions. Conversely, the upwind sites have a low percentage of NOx as NO, meaning the upwind areas are affected to a much greater degree from aged NOx emissions, perhaps attributable to earlier mobile source emissions in the local or greater Sacramento area. In both cases, though, the differences between upwind and downwind influences are dramatic.

Other analyses were conducted by aggregating data by day of week. The upwind sites show patterns that are some-what reflective of typical motor vehicle dominated condi-tions, with higher weekday levels than occur on weekends. The downwind sites, however, do not. Since the downwind sites do not show much difference by day of week, and the Yard operates seven days per week, these results are con-sistent with the influence of the rail yard at the downwind sites. The interesting aspect of the overall scope of the

data analyses is that all the results are consistent with each other, and show that the monitors are captur-ing very noticeable effects of rail yard emissions. For the 2006 summer sampling period, measurements for additional air toxic pollut-ants were included, with assistance from a EPA air toxics grant.

Community outreachTogether with Union Pa-cific, CARB and PCAPCD conducted three workshops for Roseville residents after CARB released the risk as-sessment study in October 2004. The purpose of these workshops was to commu-nicate information regard-ing potential exposure and health impacts to the local community. The workshops included an overview of the cancer risk assessments from exposure to DPM from locomotive activities at the Yard and a summary of how mathematical models are used to evaluate the potential health impacts from exposure to hazardous air pollutants. In addition to the workshops, PCAPCD staff also attended several city and neighbor-hood association meetings to provide an update for the community on the mitigation and monitoring programs at the Yard. PCAPCD received positive feedback from the attendees of these workshops and meetings. PCAPCD will continue updating the public as new information becomes available, including scheduling a tour of the air monitoring sites and the proof-of-concept demonstration project for neighborhood residents.

SummarYThis project is an example of how public and private enti-ties can work together in crafting innovative approaches to reduce community-wide health risk. The underlying key to success lies in the ability to create partnerships between regulatory agencies and the regulated community with visions for improved air quality, greater public health ben-efits, wider use of innovative technologies, and a willingness and commitment to get there.

There has been significant progress implementing the three components of the agreement since December 2004. Union Pacific, PCAPCD, and the other participating agencies are fully committed to fulfill their commitments to support both mitigation measures and the monitoring project. In addition, PCAPCD will continue to seek other opportunities to further the overall program objectives through the use of both regional incentive funds and grants.

reFerenCeS1. Roseville Rail Yard Study. Prepared

by California Air Resources Board, Sacramento, CA, October 2004.

2. Fritz, S.E. Diesel Fuel Effects on Lo-comotive Exhaust Emissions; Final Report. Prepared by the Southwest Research Institute for California Air Resources Board, Sacramento, CA, October 2000.

3. Campbell, D.E.; Fujita, E.M. Deploy-ment of the DRI Mobile Van in Sup-port of the Roseville Rail Yard Air Monitoring Project; Final Report. Prepared by Desert Research Insti-tute for Placer County Air Pollution Control District, Auburn, CA, May 2005.

4. Arnott, W.P.; Hamasha, K.; Möos-muller, H.; Sheridan, P.J.; Ogren, J.A. Toward Aerosol Light Absorp-tion Measurements with a 7-Wave-length Aethalometer: Evaluation with a Photoacoustic Instrument and 3-Wavelength Nephelometer; Aerosol. Sci. Technol. 2005, 39, 17-29.

5. Campbell, D.E.; Fujita, E.M. Data Analysis on the Roseville Rail Yard Air Monitoring Project; Interim Final Report. Prepared by Desert Research Institute for Placer Coun-ty Air Pollution Control District, Auburn, CA, June 2006.

em


feature - pubs.awma.org

Documents