• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

A Greenhouse Gas Emissions Inventory for Pennsylvania

Posted on: Tuesday, 16 August 2005, 03:00 CDT

ABSTRACT

The Pennsylvania greenhouse gas (GHG) emissions inventory presented in this paper provides detailed estimates of emissions and their sources for the six major categories of GHGs. The inventory was compiled using the current U.S. Environment Protection Agency methodology, which applies emissions factors to socioeconomic data, such as fossil energy use, vehicle miles traveled, and industrial production. The paper also contains an assessment of the methodology and suggestions for improving accounting with respect to process, sectoral, and geographic considerations. The study found that Pennsylvania emitted 77.4 million metric tons carbon equivalent of GHGs in 1990 and that this total increased by 3% to 79.8 million metric tons carbon equivalent by 1999. Despite this increase, however, the state's percentage contribution to the United States total declined during the decade. Pennsylvania's carbon dioxide (CO2) emissions from fossil fuels represented 92.4% of 1990 totals and declined to 90.5% in 1999. Electricity generation was the largest single source of CO2 emissions, being responsible for 38% of fossil fuel CO2 emissions in 1990 and 40% of the total in 1999. Transportation emissions accounted for the largest increases in emissions between 1990 and 1999, whereas industrial emissions accounted for the largest decrease. The overall trend indicates that Pennsylvania has been able to weaken the relationship between GHG emissions and economic growth.

INTRODUCTION

Most scientists believe that greenhouse gases (GHGs) are the most important contributor to long-term warming of Earth's surface.1 Most scientists also believe this warming will have negative effects on agriculture, forests, wildlife, and human health.2 Although there are some ways of reducing emissions without costs, including energy conservation, most efforts to curb GHGs will incur costs and might have wide-ranging social and economic impacts. Involvement by the U.S. government in the Kyoto Protocol to address the problem has stalled, but some states have formulated action plans to reduce GHG emissions. A GHG inventory is a necessary first step in the formulation of effective mitigation plans. Moreover, it can pay dividends in other areas because some GHGs contribute to other problems, like stratospheric ozone depletion.

Pennsylvania is one of the leading emitters of GHGs in the United States. The state ranks fourth in the nation in emitting carbon dioxide (CO2),3 the major GHG. Moreover, CO2 emissions from Pennsylvania exceed the levels from all but 12 countries.3 This article summarizes the compilation of a GHG emissions inventory for Pennsylvania. It provides estimates for 1990 and 1999 for the major categories of GHGs, as well as an interpretation of the results (1990 is the benchmark for the Kyoto Protocol, and 1999 was the most recent year for which data were available). It also performs an assessment of the limitations of the estimation methodology and suggestions for its improvements.

A GHG emissions inventory can provide valuable information to government, business, and the public. It explains how the activities of everyone in a given geographic area contribute to the generation of GHGs. It is also a necessary first step in assessing whether a state should develop a mitigation action plan, and, if so, what form the plan should take.4

The emissions inventory reports on six categories of GHGs: CO2, methane (CH^sub 4^), nitrous oxide (N^sub 2^O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), and sulfur hexafluoride (SF^sub 6^). The estimates are organized by major emissions categories, such as fossil fuel consumption; biomass fuel consumption; production processes; agriculture and livestock production; and waste disposal, treatment, and recovery. These categories are broken into subcategories in those cases when more than one GHG is emitted and when the emissions are widespread.

The study follows the U.S. Environmental Protection Agency (EPA) methodology in the Emission Inventory Improvement Program's manual, Estimating Greenhouse Gas Emissions.5 This methodology provides an organizing framework, as well as guidelines, shortcuts, and parameters for computations. Using the standardized methodology also makes it easy to compare the Pennsylvania inventory with inventories of other states. At the same time, although the standardized EPA methodology applies to all states, it does not always provide the best possible techniques for estimating Pennsylvania emissions. Consequently, we provide insights on how to enhance the EPA methodology to improve the accuracy of GHG emissions estimates for Pennsylvania. In doing so, we also comment on the role that emissions inventories can play in the formation of mitigation action plans. As our discussion will show, the methods used to estimate emissions often encourage one mitigation option over another because of the scale at which data are collected or the character of the datasets underlying emissions estimates. It is, therefore, important not only to provide the most accurate estimate of GHG emissions possible but also to provide emissions inventories that best elucidate the underlying causes of those emissions to better inform mitigation strategies.

We emphasize that the GHG estimates presented in this article are provisional. Moreover, most of the estimates are presented for only 2 years and may not be indicative of long-term trends or recent changes in those trends.

BACKGROUND

Everyday human activities can lead to climate change. Agriculture, forestry, fossil fuel production and combustion, chemical manufacture and use, and waste disposal release CO2, CH^sub 4^, N^sub 2^O, and other greenhouse gases.1 According to the consensus of scientists involved in the Intergovernmental Panel on Climate Change process, these GHGs accumulate in the atmosphere, increase radiative forcing, and in the process eventually change Earth's surface climate.1 It is important to note that atmospheric GHGs and radiative forcing are normal and necessary to life but that human activities are raising GHG concentrations and radiative forcing above natural levels. The ensuing short-term climate variability and long-term climate change have already been found to have adverse effects on flora, fauna, human health, and various natural systems. Climate change may have potentially catastrophic effects if GHG emissions are not curtailed significantly in the coming decades.2

Until the 1990s, most efforts to identify the human activities producing GHGs and to measure their emissions focused on the global level. At the global scale, the longlived GHGs considered here diffuse and mix regardless of their points of origin. This universal mixing makes it difficult to use instruments for measuring greenhouse gas emissions at subglobal scales. Instead, analysts must infer the emissions from human activity. Most countries keep broad records of land in forestry and agriculture, production and consumption of fossil fuels, chemical manufacture and use, waste disposal, and other major human activities within their boundaries. It is relatively straightforward, therefore, to construct national inventories of GHG emissions from general human activity data. The United States has used this approach to compile GHG emissions inventories since before 1990.6,7

In a large, diverse country like the United States, however, the mix of human activities and resulting GHG emissions varies from region to region. For instance, states dominated by agriculture, heavy manufacturing, or coal mining, such as Kansas, Ohio, and West Virginia, respectively, emit markedly different bundles of GHGs. If the United States were to develop a national action plan to reduce emissions but failed to account for state-by-state differences, it is unlikely that the action plan would succeed, because it would lack the detail to be cost-effective and equitable across and within regions.8,9 To develop an effective action plan for GHG mitigation, states and localities must first compile emissions inventories.4

Recognizing the need for state-level action to decrease GHG emissions, the EPA has encouraged states to compile emissions inventories for more than a decade. The state-level emissions inventory protocols4 use the international reporting standard established by the Intergovernmental Panel on Climate Change10 and expanded by the EPA.7 The GHGs cataloged by U.S. emissions inventories include CO2, CH^sub 4^, N^sub 2^O, and certain man-made gases commonly known as ozone-depleting compounds, their substitutes, and some other man-made compounds. Ozone-depleting compounds include chlorofluorocarbons (CFCs) and HFCs, which were banned under the Montreal Protocol and are no longer included in GHG emissions inventories. HFCs are the primary substitute for ozone depleting compounds; other important fluorinated compounds include PFCs and SF^sub 6^. The sectors tracked by GHG inventories include the activities associated with agricultural production, forestry, energy production and consumption, other industrial processes, and waste disposal.

Not only is there great diversity from state to state, but also there is tremendous variation within mos\t states. In Pennsylvania, various cities, counties, and regions are known for their agriculture, forestry, coal mining, transportation systems, manufacturing, or refuse disposal. Even in one small place, Centre County, PA, for example, each of these activities is important. Ultimately, substatelevel entities, such as metropolitan regions, counties, and universities, will need to compile inventories and formulate action plans.11 Several recent efforts have recognized that need. The Global Change in Local Places project12 adapted the EPA state-level methodology13 to conduct GHG inventories for select counties in North Carolina, Kansas, Ohio, and central Pennsylvania.14,8,15 The International Council for Local Environmental InitiativesCities for Climate Protection campaign independently developed tools to inventory GHGs from communities and institutions.l:l Lachman16 used this instrument to compile a GHG emissions inventory for the Penn State University Park Campus. Whereas substate inventories are important to the formulation of future GHG mitigation action plans, effective mitigation of GHG emissions can only occur through a multiscale approach; mitigation action plans must be developed at the substate, state, and national levels. Emissions inventories can inform this process by identifying the appropriate scales at which to study and mitigate different emissions sources. For instance, it may make sense to use a state- level approach for mitigating GHG emissions from electricity production, but a local scale for reducing emissions from transportation.

Methodology

The GHG emission figures presented in this report primarily were generated by using the EPA methodology for estimating GHG emissions.5 It should be noted at the outset that these figures are not based on direct measurement with the use of scientific instruments but rather are estimates of emissions based on socioeconomic data. Therefore, our GHG emissions estimates are only as accurate as the underlying socioeconomic data and the accounting practices, transformations, and emissions and conversion factors applied to those data. Details of the methods used to calculate GHG emissions estimates from each sector are presented in our report to the Commonwealth of Pennsylvania.17 The emissions were estimated through the use of standard emission coefficients applied to socioeconomic data, such as energy use, industrial activity, and vehicle miles traveled (VMT).5

We work through an example here to demonstrate the process of estimating emissions from socioeconomic data. In the case of municipal waste management, GHG emissions and sequestration have three sources: CH^sub 4^ emitted from waste at each landfill; CO2 and N^sub 2^O emitted from the annual combustion of municipal solid waste; and carbon sequestered by the annual addition of waste to landfills. Calculation of emissions from combustion of municipal solid waste and carbon sequestration from landfilling are relatively simple, because annual data on waste combustion and landfilling are available (according to Emission Inventory Improvement Program, some organic material does not escape landfills in gaseous form5). Multiplying these quantities by coefficients that reflect the average emissions and sequestration per ton of waste provides reliable estimates of the emissions and sequestration resulting from each activity.

CH^sub 4^ emissions from municipal solid waste are less straightforward. Once placed in a landfill, municipal solid waste can emit CH4^sub ^ for as long as 30 years.5 Thus, to calculate CH^sub 4^ emissions, it is necessary to know how much waste is in place at each landfill in the state and the age of that waste; however, these data are not readily available in Pennsylvania. In this case, the EPA methodology provides a formula for estimating waste in place using gross population data, the average waste disposal per capita, and a 30-yr multiplier. Although this methodology is imprecise, the resulting figure provides a reasonable approximation of CH^sub 4^ emissions that does not require costly data collection from each landfill dating from 1960. Still, a major disadvantage to this approach is that it offers few mitigation options other than simply to reduce the average amount of waste generated per capita, whereas more detailed analyses might suggest other options. For instance, more detailed information about each landfill in question would suggest other options, such as flaring of gases or different waste-management techniques. We address this issue in a later section of this article.

Throughout this inventory, with one exception, we have estimated Pennsylvania GHG emissions using the EPA methodology. In some cases, the estimates are the best available; in other cases, we believe the estimates for Pennsylvania could be improved with marginal increases in time and money. Whereas improving the accuracy of the methodology used to estimate Pennsylvania GHGs is important, it is equally important to generate estimates based on the EPA methodology, despite any flaws or shortcomings. Assuming all 50 states eventually complete inventories of GHG emissions, those inventories must be directly comparable, not only with each other, but with national- scale inventories. Consistency in methodology is critical to achieving this goal. Toward that goal, this article summarizes our major findings based on the EPA methodology.

RESULTS

The results of the Pennsylvania inventory are presented in Tables 1 through 11, which are distinguished by category of GHG and emissions source. Descriptions of the processes by which the GHGs are emitted and the data and formulas used to calculate the emissions, as well as spreadsheets of the actual calculations are presented in Rose et al.17 Throughout this report, emissions of all GHGs are expressed in metric tons of carbon equivalent (MICE) or million metric tons of carbon equivalents (MMTCE). This is a common denominator for the different radiative forcings, or warming potentials, of the various GHGs.

Table 1. Summary of all PA GHG emissions (MMTCE).

A summary of Pennsylvania GHG emissions is provided in Table 1 and in Figures 1 and 2. The table indicates that CO2 emissions from fossil fuels clearly dominate the picture. In 1999, for example, this category contributed 72.23 MMTCE of the total of 79.79 MMTCE generated in Pennsylvania, or 90.53% of the total. Because of data limitations, we were not able to accurately describe emissions within Pennsylvania from mobile sources associated with interstate and international transportation, such as marine vessels, railroads, and aircraft. We have computed emissions from fuels purchased in Pennsylvania for these transport modes, but this calculation is unlikely to correspond to GHGs actually emitted within the state's boundaries. This problem of measuring transportation sector emissions also applies to motor vehicles from large trucks involved in interstate commerce and to personal automobiles driven by commuters living near state boundaries.

Table 2. PA CO2 emissions from fossil fuel combustion (MMTCE).

Figure 1. Relative contribution of each sector to total Pennsylvania GHG emissions, 1990.

Several other emission categories besides fossil fuels are significant (Table 1). Note also that forestry and landuse change has negative net additions to CO2 emissions. This negative value reflects the fact that plants absorb CO2 from the atmosphere during photosynthesis and that additional forest growth sequesters even more of this GHG.

To put the totals from Table 1 into perspective, Pennsylvania's population was 4.8% of the United States total in 1990 and 4.4% of the total in 1999. Using the data to calculate carbon equivalent GHG emissions in Table 1, Pennsylvania's contribution to the national total of GHGs is 5.6% and 4.9% for 1990 and 1999, respectively. Thus, despite an increase in the overall level of GHG emissions in Pennsylvania between 1990 and 1999, we see a downward trend in the state's share of the national total of GHGs being emitted. Furthermore, whereas Pennsylvania emitted more GHGs per capita than the national average in both years, per capita emissions for the state declined.

None of the categories of GHGs, other than CO2 from fossil fuel combustion, generates >5% of total emissions. The most notable of these minor sources are GHGs from nonenergy industrial sources (all six categories combined); CH4 leakages from oil and natural gas extraction, transportation, and storage; CH^sub 4^ releases from coal mining; and GHG emission (primarily CH^sub 4^) from municipal waste management and from manure management. All of the other noncarbon emissions amount to -1% or less of the Pennsylvania total emissions.

Figure 2. Relative contribution of each sector to total Pennsylvania GHG emissions, 1999.

Figure 3. Sectoral contributions to Pennsylvania CO2 emissions from fossil fuel combustion, 1990 and 1999.

A breakdown of CO2 emissions from fossil fuel combustion by customer (Table 2 and Figure 3) indicates that electricity generation is the dominant source of this GHG. Tables 3-11 provide similar disaggregations by emission types and sources. For example, Table 4 provides disaggregations of GHG emissions from nonenergy industrial processes by gas and by source. This category increased more than any other, >50% between 1990 and 1999. The most prominent contributors to this total are CO2 from cement manufacture and HFCs and PFCs. None of the other sources accounted for >10% of the 1999 emissions from the nonenergy industrial processes listed in Table 1.

Table 3. Extended presentation of PA CO2: emissions by sector and fuel type (MMTCE).

Table 4. PA GHG emissions from nonenergy: industrial processes, by activity (MTCE).

Caveats and Methodological Shortcomings

Note that our estimates are based on the latest data from the Energy Information Agency (EIA) on Pennsylvania coal use (unpublished data rele\ased in late October 2002).16 EIA implemented major modifications in its methodology, primarily to account for the reclassification of many electric utilities as nonutility generators accompanying electricity industry deregulation. We have distinguished these two subcategories of electricity generation in Table 12.18-19 Note that the revised coal use estimate for electricity generation for 1999 is significantly higher than the previously published standard data,19 which failed to include nonutility generation. The revised EIA data are similar to those published later by EIA.20

Table 5. PA GHG emissions from oil and natural gas (MICE).

A major caveat relates to the transportation sector, which is experiencing significant growth in CO2 emissions. Motor vehicle traffic generates a large proportion of total transportation emissions. Emissions were calculated for motor vehicle traffic by following the EPA methodology and using EIA data detailing motor fuel sales in Pennsylvania. Although the report shows nearly 17% growth in CO2 emissions from motor vehicles from 1990 to 1999, examination of more recent trends indicates that emissions from motor vehicles may be rising at an even faster rate. For instance, annual VMT in Pennsylvania grew by 20% from 1990 to 1999.21-22 Whereas the average efficiency of the private fleet improved marginally during that period (~1.8%23), fuel efficiency standards for new light-duty cars and trucks declined nationally by -4% over the same period. Furthermore, consumers demonstrated a growing preference for purchasing sport-utility vehicles over more fuel- efficient vehicles during the 1990s. Whereas average motor vehicle efficiency improved consistently from 1976 to 1992, it declined by ~0.2 miles per gallon in 1993 and has fluctuated around 16.7 miles per gallon since then, showing only marginal increases or decreases from year to year.27 All of these factors may have contributed to a higher rate in CO2 emissions growth in the latter half of the 1990s and the first few years of 200Os than the rate described in this report for 1990-1999. Transportation emissions will need to be investigated further in any future analysis, and the methodology and growth rates in this report will need to be improved to generate a more accurate estimate of emissions from the transportation sector.

Several other shortcomings arise in the EPA methodology. For example, nitrous oxide generated from nitric acid production for Pennsylvania is calculated as a percentage of U.S. production on the basis of Pennsylvania's share of the U.S. total population. Thus, even though Pennsylvania's nitric acid production capacity was the same in 1999 as in 1990, it is estimated that the generation of nitrous oxide from nitric acid production in the state declined from 107.8 thousand MTCE to 101.3 thousand MTCE over the decade because Pennsylvania's population share decreased (see Table 4). This case points out the limitations of using population ratios to estimate greenhouse gas emissions. Similar problems will likely be found under GHGs from nonenergy industrial sources.

Table 6. PA GHG emissions from municipal waste management (MTCE).

Table 7. PA GHG emissions from manure management (MTCE).

The EPA methodology also contains a potentially important omission relating to CFCs. CFCs are greenhouse gases but were not included in the Kyoto Protocol. Because the earlier Montreal Protocol banned them, framers of the Kyoto protocol assumed that these gases would be phased out before the compliance period of 2008- 2012. Furthermore, there is considerable uncertainty in the factors used to convert CFCs and HFCs to MTCE measurements. This uncertainty has led to the omission of such calculations in the U.S. national inventories, although the United States has expressed an interest in including these emissions estimates in the reporting documents. Because the United Nations Framework Convention on Climate Change, Intergovernmental Panel on Climate Change, and United States do not include CFCs and hydrochlorofluorocarbons in their inventories and projections, they are not considered in our inventory for Pennsylvania. However, ozone-depleting compounds substitutes (e.g., HFCs) and some other fluorinated compounds (e.g., PFCs and SF6) are in fact GHGs and are included in the Kyoto Protocol. Emissions of these gases from Pennsylvania increased from 14.3 thousand MTCE in 1990 to 814.8 thousand MTCE in 1999.

Several important causal trends can be identified despite these caveats; that is, these trends hold generally over the course of the 1990s even if we take into account data omissions. Recent studies have identified underlying "sources" of change in GHGs (e.g., economic growth, population growth, change in fuel mix, and change in a sectoral intensity of total economic activity) and have used a formal methodology called structural decomposition analysis to identify their relative contributions over time.24,25 Here we use this concept to isolate major explanatory factors and discuss their influence on a less formal basis.

Table 8. PA GHG emissions from burning of agricultural waste (MTCE).

Table 9. PA GHG emissions from municipal waste water (MTCE).

For example, had only economic growth taken place during the 1990s, with no changes in any of the other underlying causal factors, overall CO2 emissions would have grown at a commensurate rate of 24%. Yet, overall CO2 emissions have increased only slightly over the decade. The good news is that in Pennsylvania, there has been a weakening of the relationship between certain types of economic activity (and therefore GHG emissions) and economic growth. Many analysts were previously convinced that electricity use had to move in lockstep with economic activity, but this conclusion is clearly no longer valid, because other factors have offset the upward pressure of economic growth on electricity use and, hence, on CO2 emissions. Some of the causal factors exerting downward pressure on emissions include a relative decrease in manufacturing from 21% of gross state product in 1990 to 19% in 1999. In addition, there was an 18.8% increase in fossil energy conservation in this sector (11.2 BTUs per dollar of output in 1990 and 9.1 BTUs per dollar of output in 1999, both expressed in 1999 constant dollars).

Potential Methodological Improvements

Although the EPA methodology is the most widely accepted methodology in the United States for generating GHG emissions inventories, it is not without problems and can and should be improved. This section outlines some significant improvements that should be included in future inventories to improve the accuracy and precision of emissions estimates. It is important to note that the improvements suggested here are not substitutes for the current EPA methodology but are suggested as additional calculations. Because one of the primary goals of any state-level GHG inventory is compatibility and comparability with other state inventories, a common methodology must continue to be used. Nonetheless, the EPA methodology itself encourages states to perform additional calculations where improvements in emissions estimates are possible, and these additional calculations, and the methods behind them, should be included as part of future emissions inventories to ensure that the estimates reported are the most accurate possible.

Table 10. PA GHG emissions from mobile combustion (MTCE).

Table 11. PA GHG emissions from stationary combustion (MTCE).

Of course, this discussion would be moot if it were possible simply to take atmospheric measurements at the point of release of all GHGs and to report those emissions in a disaggregated manner, across various socioeconomic sectors. Unfortunately, establishing a scientifically sound GHG monitoring network at every source of emission would be financially prohibitive, if not physically unfeasible. Other factors also preclude this approach. Most important, there are many significant nonpoint sources of GHG emissions, including agriculture, land-use change, and transportation. Because CO2 and other GHGs mix with other atmospheric gases quickly and evenly, it is impossible to use atmospheric monitoring devices, such as the devices used to measure global CO2 concentrations, to perform local assessments.

Furthermore, a major justification for performing GHG emissions inventories at scales smaller than the national or even international scales is to identify socioeconomic activities that are responsible for GHG emissions to formulate GHG mitigation plans. In the absence of a scientifically sound emissions monitoring network and in the absence of the possibility of creating such a network, GHG inventories have focused on identifying socioeconomic processes that both are responsible for GHG emissions and are measurable and then estimating GHG emissions from those processes based on our understanding of the chemistry that ultimately results in emissions.

Table 12. Fuel utilization in electricity in Pennsylvania, 1990- 1999 (in trillion BTU).

Issues relating to the accuracy and precision of estimates are primarily accounting issues and can be classified into three categories: sectoral accounting issues, geographic accounting issues, and process-based accounting issues. sectoral accounting issues refer to the problem of clearly identifying which economic sector is responsible for any given emissions estimate. Geographic accounting issues refer to the problem of precisely and accurately attributing emissions to a specific geographic entity, in this case Pennsylvania. Process-based accounting issues are a bit more complicated but constitute a critical consideration in the development and improvement of emissions inventories.

As previously stated, one of the major justifications for local- and state-based GHG inventories is to identify place-specific estimates of GHG emissions, by socioeconomic sector or process, to aidresearchers, policy-makers, and stakeholders who will use the inventory to generate action plans aimed at reduction of GHG emissions. In addition, GHG emission inventories are necessarily based on our understanding of the socioeconomic processes and the resulting chemical processes that ultimately result in GHG emissions. Whereas our calculations of GHG emissions from a given sector may be based on large-scale data (e.g., in the case of transportation, gasoline sales), the actual processes resulting in those emissions may be, and often are, smaller in scale. Extending the case of transportation, although gasoline sales may provide a reasonable estimate of GHG emissions at a state level, attributing this as the proximate cause of GHG emissions from transportation can be misleading to policy-makers because it may suggest that the only way to reduce GHG emissions is to use a top-down approach aimed at reducing gasoline sales, such as tax incentives.

In reality, the decisions that result in GHG emissions from transportation occur at multiple scales. To be sure, a national process that involves gasoline prices and the "invisible hand" is to some degree implicated in the GHG emissions. Various econometric analyses have shown that as gasoline price increases, efficiency also increases, resulting in lower gasoline consumption. However, large-scale processes are only part of the processes involved in GHG emissions from transportation. Other factors, such as the origin and destination of a trip, the transportation network available, and the mode of transportation chosen, are decided by individual travelers on a much smaller scale. These processes are entirely ignored by the current EPA methodology because the focus is simply on estimating the weight of CO2 and other GHGs, and large-scale data are sufficient to achieve this goal. If the goal of fully informing policy-makers and stakeholders about the sum of processes responsible for GHG emissions is to be met, however, steps should be taken, when possible, to provide richer and more detailed descriptions of these processes, both at large and small scales.

All three of the accounting issues are critical to a more balanced, accurate, and informative GHG emissions inventory. Without accurate and precise geographic accounting, emissions estimates may not reflect the true emissions of a state. Without proper sectoral accounting, policy-makers may misallocate resources and energy into emissions reduction plans that do not address the true causes of emissions. Finally, without sufficiently detailed descriptions of the multiscale processes that are proximate causes of emission patterns, policy-makers cannot approach the problem of mitigating GHG emissions with a sufficiently multitiered and locally specific plan, which is one of the major justifications for performing emissions inventories in the first place. The following section of this document explores flaws and potential improvements to the GHG inventory methodology. These improvements are organized into three sectors: transportation, land-use change, and electricity generation. In addition, improvements based on more accurate and precise local data are explored with regard to emissions from land- use change and municipal waste management.

Methodological and Accounting Issues

Accurately attributing emissions to specific geographic entities is the primary goal of and justification for performing state-level GHG inventories. The major failing of the EPA methodology in this area is attributable primarily to the oversight of bunker fuels. In the context of GHG emissions inventories, bunker fuels are fuels that are purchased in one state but actually consumed in another. One example of bunker fuels offered by the EPA is that of jet fuel.5 Under the current methodology, jet fuel emissions are calculated based on state fuel sales. However, most jet traffic leaves the state after fueling and takeoff and, therefore, the actual physical emission of GHGs occurs outside the state boundaries. Similarly, emissions from interstate jet traffic entering Pennsylvania are attributed to the origin state, rather than Pennsylvania, where the emissions actually occur. The EPA methodology suggests ignoring bunker fuels when performing state-level inventories because of the complexity in calculating them, and in the case of emissions from the combustion of jet fuel, the benefits of accurately characterizing jet fuel and aviation gasoline bunkers may indeed be outweighed by costs of collecting and processing the data because jet and aviation fuel combustion was responsible for only 3.5% of 1990 estimated emissions.

Transportation. Motor gasoline consumption was responsible for >13.8% of total estimated emissions for 1990, however, and this estimate suffers from the same problem of geographic accounting. Emissions estimates for consumption of motor gasoline are based on total gasoline sales in Pennsylvania, although Pennsylvania has considerable interstate traffic, serving as one of the major interstate corridors to and from the eastern seaboard. Unlike jet fuel, bunkers for motor gasoline can be calculated with minimal time and monetary resources using readily available estimates of VMT from the Office of Highway Information Management. In fact, emissions estimates of nonCO2 GHGs from mobile combustion of fossil fuels use these VMT estimates. Furthermore, the Pennsylvania Department of Transportation has estimates of VMT for the state. We are currently exploring available data to determine the best source of VMT data, by vehicle type, to characterize actual CO2 emissions from motor gasoline more accurately.

Whereas this method may provide better estimates of GHG emissions resulting from transportation, it does little to improve policy- makers' understanding of the socioeconomic processes that generate the traffic being measured. Using local-scale data available from the U.S. Census, however, it is possible to explore these processes at a more local and detailed scale. Work by Neff26 uses census data describing commuters' origins, destinations, and mode of transportation to generate spatially specific characterizations of GHG emissions resulting from the commute to and from work, which is responsible for roughly one third of national GHG emissions from transportation. Attributing emissions to the places from which they are emitted allows a more detailed examination of smaller-scale processes, such as urban sprawl, accessibility to public transportation, and public transit ridership. Nonetheless, it is important to note that this approach requires very detailed origin- destination data, disaggregated by mode of transportation. Whereas such data are available for commuters in every metropolitan area in the United States, they are not available for commercial traffic, nor are they available for personal trips, such as running errands and visiting family and friends, or for extended trips from one locality to another.

Nevertheless, locally specific examinations of the processes resulting in GHG emissions are critical to public policy aimed at mitigating these emissions and should be included as supplementary information, where monetary and temporal constraints on the inventory process permit. Whereas it is not the role of a GHG inventory to suggest mitigation options, supplying policy-makers, stakeholders, and researchers with the necessary information to formulate such a plan is a critical role of all inventories. By providing smaller-scale analyses of the processes implicated in GHG emissions, researchers would provide those involved in formulating action plans valuable information that may suggest locally specific approaches to GHG mitigation that may otherwise not be considered.

Land-Use Change. Another area where an understanding of smaller- scale processes is critical to understanding GHG emissions is that of land-use change. As in the case of transportation, the estimation of emissions from landuse change is based on state-level data. In this case, those data consist of forest composition and scope and of changes in those forests resulting from activities such as logging, conversion to agricultural land, and encroachment on forested land by the expansion of urban and suburban communities. As in the case of transportation, the description of the actual processes resulting in these changes is lacking from the current methodology and, again, like transportation, a richer and more detailed exploration of these processes is possible through greater attention to the geography of land-use change.

Current estimates of land-use change emissions are based on state- level data gathered by the U.S. Forest Service at 5-year intervals and extrapolated using temporal trends. Smaller-scale and more- current data on landuse patterns are available through remote sensing. Using satellite data on land use is not entirely straightforward and requires significant expertise, particularly in classifying visual images so the resulting data accurately reflect the composition of land cover observed on the ground. Nevertheless, such classification is possible, and the resulting spatial data could be used to achieve both more accurate accounting of the spatial extent and content of forested areas and more spatially specific accounting of the change, allowing researchers to perform local assessments of the locally specific processes resulting in that change.

Electricity Generation. Another major area of concern is electricity generation. Here, the problem is multifaceted and suffers from issues of both geographic and sectoral accounting. Geographically, whereas emissions can be easily identified as physically originating in one state or another, the actual use of the electricity generated is difficult to attribute to any specific state. Such attribution is critical for two reasons. First, Pennsylvania is a major exporter of electricity. The\refore, the electricity produced in Pennsylvania physically emits GHGs in the state, but the actual socioeconomic processes responsible for a significant portion of those emissions occur outside the state and outside the control of Pennsylvanian policy-makers. second, the practice of wheeling makes it difficult to determine the exact fuel mix used to generate electricity consumed within the state. Wheeling is a practice allowing consumers to purchase electricity that is actually generated far from where it is consumed. For instance, Pennsylvanians have the option to purchase electricity that is generated outside the state using renewable energy sources such as wind and hydroelectric dams. Emissions estimates that assume Pennsylvania consumers use only coal-intensive electricity produced within the state may overestimate emissions.

A second flaw in the current methodology for estimating emissions from electricity generation is one of sectoral accounting. The EIA recently made significant adjustments in data reporting for electricity generation and associated CO2 emissions by nonutility generators. The change is an outgrowth of the strong effect of electricity industry deregulation (or restructuring). Many major utilities have sold or spun off their CO2-emitting generation facilities (power plants) to unregulated subsidiaries; the major utilities now only operate transmission or distribution systems, which do not emit CO2. Many of the power plants that have changed hands are now listed under the nonutility generators category rather than under electric utility generation.20

The restructuring should have resulted in data compilations that showed sizable decreases in electric utility emissions but only modest increases in nonutility generation,18 which would have been grouped into the "industrial" emissions category before 1999.27 However, the decrease in the former category is simply because of an accounting shift rather than any actual efforts at reducing emissions (e.g., in Pennsylvania, there has been relatively little increase in natural gas displacing coal as a fuel source in recent years, in contrast to the relatively large gas for coal displacement nationwide). More recently, an effort has been made to account for all utilities that were reclassified, resulting in a much larger estimate of nonutility power producer activity in 1999, as presented in Table 12.18

In addition, EIA has gone to a dual reporting format for electricity-related CO2 emissions. Rather than only presenting electricity generation and emissions on the production (supply) side, EIA now reports emissions according to the end user (or demand) side as well. Although not explicitly stated as a purpose for this change, one implication is to take the onus of CO2 emission creation off the shoulders of the producers alone and to suggest that consumers, through their purchases of electricity, are also responsible for these emissions.

On the supply side, emissions are divided into five categories: industrial, commercial, transportation, residential, and electricity utilities. On the demand side, however, there are only four categories; the utility sector emissions are "shared out" to the four other end user sectors (i.e., industrial, commercial, transportation, and residential) according to their purchases of electricity.

The shift is a positive step toward evaluating the impetus for GHG emissions and for developing mitigation strategies that are applicable to both the demand and supply sides. Some economists prefer mitigation measures that are most "cost-effective" regardless of on which side they are imposed; others include the concept of "responsibility" for emissions in formulating policy, including the use of transfers, such as taxes or subsidies, to adjust for this concept.

Nonetheless, the new dual reporting format falls short of the more comprehensive adjustments that may be applicable. For example, EIA now only specifies enduse emissions at the national level, and the analyst must adjust the data to downscale them to the state level. Unfortunately, data may only be available to specify use within each state on a "net" basis. That is, a factory may purchase electricity from a generator outside the state, yet the basic data characterize emissions associated with this electricity as "within- state" emissions. Likewise, a generator in the state may sell to users outside the state, but the sale cannot be separated from in- state sales. Effectively, interstate sales (gross flows) are omitted, and only a net balance can be inferred. A significant data analysis (probably on a plant-by-plant basis) for electricity produced within the state and for electricity imported into the state is required for a truly accurate assessment.

Municipal Waste Management. Finally, improvements based on more locally specific data can be applied and, in fact, are being applied to estimates of GHG emissions from municipal waste management. As previously mentioned, lack of statewide data describing total waste in place at Pennsylvania landfills necessitated using national and regional averages and trends to estimate methane emissions from landfills. As the EPA continues to release detailed spreadsheets to aid researchers with state-level inventories, improvements in the methodology that take advantage of more readily available smaller- scale data are being incorporated in those spreadsheets. In the case of municipal waste management, the spreadsheets provide a method for estimating waste in place based on historical data describing the annual amount of waste dumped at landfills. This methodology was not previously described in the official methodology and constitutes a significant improvement in the accuracy of GHG emissions estimates from landfills. Thus, it is important for researchers to update their methodology continually using improvements made by others, particularly the EPA, which sets the standards for state-level inventories. A clearinghouse or forum that provides these improvements to the user community would be extremely valuable.

CONCLUSIONS

To summarize our results, Pennsylvania's total GHG emissions have increased 3%, from 77.44 MMTCE in 1990 to 79.79 MMTCE in 1999. However, Pennsylvania's contribution to the national GHG emissions total has decreased from 5.6% to 4.9% for the years 1990 and 1999, respectively. Whereas Pennsylvania's share of the national total of GHG emissions has been declining, its contribution is still above the national average on a per capita basis. In 1999, Pennsylvania emissions were 6.46 MTCE per person; the United States per person average for 1999 was 6.10 MTCE. Furthermore, in economic terms, Pennsylvania's 1999 emissions represented 0.47 Ib of carbon equivalent per dollar of state economic output compared with the national average of 0.40 Ib.

The inventory indicates that CO2 emissions from fossil fuels are, by far, the most significant GHG. In 1999, this category contributed 72.23 MMTCE of the total 79.79 MMTCE generated in Pennsylvania or 90.53% of the total. Despite their large contribution to the total, GHGs from fossil fuels increased <1% between 1990 and 1999. Of the fossil fuels consumed, bituminous coal is the largest fuel category contributing to Pennsylvania's GHG emissions. In 1999, bituminous coal use among all sectors contributed 30.62 MMTCE or 38% of all GHGs reported in the inventory.

The inventory also demonstrates that motor vehicles are the fastest growing source of GHGs in Pennsylvania. Although the state's population is essentially static, total vehicle miles are rising, which, in turn, means that per capita VMT are rising.21,22 This fact, in conjunction with consumer preferences for larger, less efficient vehicles, is primarily responsible for the rise in total GHG emissions from the transportation sector, both in Pennsylvania and in the United States.23

The Pennsylvania GHG emissions inventory highlights other activities that produce GHGs. For instance, the state has plentiful, relatively pure limestone available for lime calcining and cement manufacturing-production processes that emit disproportionately high quantities of CO2. Still, the dominance of fossil fuel use, especially for energy production and transportation, overwhelms all of the other sources of GHGs in Pennsylvania. In light of this fact, the importance of advancing techniques to understand the multiscale processes responsible for fossil-fuel GHG emissions cannot be overstated.

This brief overview of potential and existing improvements to the EPA methodology has demonstrated several approaches that are available for improving the accuracy and precision of GHG emissions estimates, as well as for improving the understanding of the multiscale processes that are the proximate causes of GHG emissions. Both considerations are fundamental to comprehending GHG emissions and are critical inputs into the logical extension of GHG inventories-the use of this information by policy-makers, decision- makers, and other stakeholders in formulating GHG mitigation action plans. The potential improvements described here are not trivial, neither in their implications for future inventories nor in their increased complexity. As states across the nation continue to implement and improve their GHG inventory and mitigation programs, they will continue to balance the need to describe the physical emissions and their causes with the monetary, temporal, and intellectual resources required to achieve accuracy. Nonetheless, if we are to continue to improve our understanding of the state- and local-scale components of this national and global concern, three issues must be addressed: accurate geographic accounting of emissions, accurate attribution of these emissions to the correct socioeconomic sectors, and a richer and more detailed understanding of the local-scale processes that contribute to GHG emissions. In addition to issues t\hat we have covered, we also recommend consideration of other improvements, such as developing the ability to compile inventories in near-real time. Nonetheless, our results demonstrate that the EPA methodology does generate reasonable GHG emissions estimates.

ACKNOWLEDGMENTS

The research was a project of the Pennsylvania Consortium for Interdisciplinary Environmental Policy Committee on Energy and Climate Change, with the support of an External Advisory Panel comprised of members of government, industry, and other nongovernmental organizations. The authors acknowledge funding for research done in connection with this Report from the Pennsylvania Department of Environmental Protection (Contract FRL-6964-2). The authors thank the following people for their contributions, data, assistance and/or comments: Mark Linstedt (Pennsylvania Agricultural Statistics Service); Richard Birdsey (U.S. Forestry Service); Clark Talkington (Coalbed Methane Outreach Program, EPA); Scott Bartos (Global Programs Division, EPA); Laurel Belding (Communication Department, PaCBI); Thomas Kevin Swift (American Chemistry Council); Carolyn Read (Directory of Chemical Producers Program); and Julia Hutchins (EIA). The authors also acknowledge the helpful comments and other assistance of Don Brown, Patrick McDonnell, Joseph Sherrick, and Chris Trostle (Pennsylvania Department of Environmental Protection). Mohammed Kharbach and Cheng-Hau Peng served as graduate research assistants on the project. The views expressed in this report, however, are those of the authors and do not necessarily represent the position of any of the persons or organizations named above. The authors are solely responsible for any errors and omissions.

IMPLICATIONS

Action on assessing the human contribution to climate change has recently shifted from the federal to the state and local levels. A greenhouse gas emissions inventory is a necessary step in helping business leaders decide on voluntary action and in providing a basis for legislation to mandate mitigation. This paper provides insight into the situation for Pennsylvania over the past decade. Our suggested refined methodology can serve as a model for other states and localities.

REFERENCES

1. Houghton, J. T.; Ding, Y.; Griggs, DJ.; Noguer, M.; van der Linden, P. J.; Xiaosu, D., Eds. Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC). Cambridge University Press: Cambridge, UK, 2001.

2. McCarthy, J. J.; Canziani, O.P.; Leary, N.A.; Dokken, DJ.; White, K.S., Eds. Climate Change 2001: Impacts, Adaptation & Vulnerability. Contribution of Working Group Il to the Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC). Cambridge University Press: Cambridge, UK, 2001.

3. State and Local Climate Change Policy Actions. Center for Clean Air Policy: Washington, DC. 2002.

4. Rose, A. Greenhouse Gas Mitigation Plan: An Overview. Penn State/ Dickinson Environmental Law Review 2003, 12, 153-71.

5. Estimating Greenhouse Gas Emissions; EIIP Documentation Series, Volume VIII; Washington, DC: Greenhouse Gas Committee, Emission Inventory Improvement Program, and US Environmental Protection Agency, 1999. Available at: http://www.epa.gov/ttn/chief/ eiip/techreport/volume08/index.html (accessed June 20, 2005).

6. Emissions of Greenhouse Gases in the United States 2000; Report No. EIA/DOE 0573(2000); Washington, DC: Energy Information Administration, U.S. Department of Energy, 2001. Available at: http:/ /www.eia. doe.gov/oiaf/1605/1605a.html#links (accessed June 20, 2005).

7. U.S. Inventory of Greenhouse Gas Emissions and Sinks, 1990- 1999. Report No. EPA 236-R-01-001; Washington, DC: Office of Atmospheric Programs, U.S. Environmental Protection Agency, 2001. Available at: http:// yosemite.epa.gov/globalwarming.nsf/content/ ResourceCenterPublicationsGHGEmissionsUSEmissionsInventory2001.html (accessed June 28, 2005).

8. Global Change in Local Places: Estimating, Understanding, and Reducing Greenhouse Gases; Association of American Geographers Global Change in Local Places Research Team, Eds. Cambridge University Press: Cambridge, UK, 2003.

9. Rose, A.; Zhang, Z.X. Intraregional Burden-Sharing of Greenhouse gas Mitigation in the United States; Mitigation and Adaptation Strategies for Global Change 2004, 9, 477-500.

10. Revised 1996IPCC Guidelines for National Greenhouse Gas Inventories (3 Volumes). Kanagawa, Japan: Technical Support Unit at the Institute for Global Environmental Strategies, National Greenhouse Gas Inventories Programme, Intergovernmental Panel on Climate Change, 1997. Available at: http://www.ipcc- nggip.iges.or.jp/public/gl/ invsl.htm (accessed June 20, 2005).

11. Wilbanks, T.; Kates, R. Global Change in Local Places: How Scale Matters; Climatic Change. 1999, 43, 601-628.

12. Kates, R.; Torrie, R. Global Change in Local Places; Environment 1998, 40, 5 and 39-41.

13. Kates, R.; Mayfield, M.; Torrie, R.; Witcher, B. Methods for Estimating Greenhouse Gases From Local Places; Local Environ. 1998, 3, 279-297.

14. Easterling, W.; Polsky, C.; Muraco, W.; Goodin, D.; Mayfield, M.; Yarnal, B;. Changing Places, Changing Emissions: the Cross- Scale Reliability of Greenhouse gas Emissions in the U.S.; Local Environ. 1998, 3, 249-264.

15. Denny, A.; Yarnal, B.; Polsky, C.; Lachman, S. Global Change and Central Pennsylvania: Local Resources and Impacts of Mitigation. In Global Change in Local Places: Estimating, Understanding, and Reducing Greenhouse Gases; Association of American Geographers Global Change in Local Places Research Team, Eds; Cambridge University Press: Cambridge, UK, 2003; pp 122-140.

16. Lachman, S. A Greenhouse Gas Inventory of the University Park Campus of the Pennsylvania State University; Unpublished M.S. Thesis, Department of Geography, Pennsylvania State University, University Park, PA, 1999.

17. Rose, A.; Yarnal, B.; Neff, R.; Greenberg. H. Greenhouse Gas Emissions Inventory for Pennsylvania. Report to Pennsylvania Department of Environmental Protection; Center for Integrated Regional Assessment, Pennsylvania State University: University Park, PA, 2003.

18. Energy Information Administration. Revised Energy Consumption Data for Pennsylvania (unpublished); Energy Information Administration, U.S. Department of Energy, 2002.

19. Energy Information Administration. State Energy Data Report, 1999; Washington, DC: Energy Information Administration, U.S. Department of Energy, Report No. EIA/DOE, 2001. Available at: http:/ /www. eia.doe.gov/emev/sedr/contents.html (accessed June 20, 2005).

20. Energy Information Administration. State Electricity Profiles (Pennsylvania). Washington, DC: Energy Information Administration, U.S. Department of Energy, 2002. http://www.eia.doe.gov/cneaf/ electricity/ st_profiles/pennsylvania/pa.html (accessed June 20, 2005).

21. Office of Highway Information Management. Highway Statistics 1994; Washington, DC: Office of Highway Information Management, Department of Transportation, 1994. Available at: http:// www.fhwa.dot. gov/ohim/ohimstat.htm (accessed June 20, 2005).

22. Office of Highway Information Management. Highway Statistics 1999; Washington, DC: Office of Highway Information Management, Department of Transportation, 1999. Available at: http:// www.fhwa.dot. gov/ohim/ohimstat.htm (accessed June 20, 2005).

23. Energy Information Administration. Annual Energy Review 2002. Washington, DC: Energy Information Administration, U.S. Department of Energy, Report No. DOE/EIA-0384(2002), 2003. Available at: http:/ /www.eia.doe.gov/emeu/aer/contents.html (accessed June 20, 2005).

24. Casier, S.; Rose, A. Carbon Dioxide Emissions in the U.S. Economy; Environ. Resource Economics 1998, 11, 349-63.

25. Zhang, F.; Ang, B. Methodological Issues in Cross/Country/ Region Decomposition of Energy and Environmental Indicators; Energy Economics 2001, 23, 179-90.

26. Neff, R. Transportation, Urban Development, and Greenhouse Gases: Patterns of Consumption and Justice in Philadelphia, PA. Unpublished Ph.D. Thesis, Department of Geography, Pennsylvania State University, University Park, PA, 2005.

27. Electric Power Annual 2000, Vol. 1; Washington, DC: Energy Information Administration, U.S. Department of Energy., 2001b. Available at: http://www.eia.doe.gov/cneaf/electricity/epavl/ epavl_sum.html (accessed June 20, 2005).

Adam Rose, Rob Neff, Brent Yarnal, and Howard Greenberg

Department of Geography and Center for Integrated Regional Assessment, Pennsylvania State University, University Park, PA

About the Authors

Adam Rose and Brent Yarnal are professors of geography at The Pennsylvania State University. Rob Neff is assistant professor of geography and environmental systems at the University of Maryland, Baltimore County. Howard Greenberg is a senior scientist at The Pennsylvania State University. Address correspondence to: Adam Rose at the Department of Geography, 302 Walker Building, The Pennsylvania State University, University Park, PA 16802. Phone +1- 814-863-0179; Fax: +1-814-863-7943; e-mail: azr1@psu.edu.

Copyright Air and Waste Management Association Aug 2005

Source: Journal of the Air & Waste Management Association

More News in this Category