Inventory Case 9: Oregon DEQ GHG Baseline Forecast (2004)

Responsible Party: 
Oregon DEQ, in support of the Oregon Governor’s Advisory Group on Global Warming
Case Study Year: 
2004
State Involved: 
Oregon
Country Involved: 
United States
Case Study Type: 
Inventory Case

1. Who did it: Oregon DEQ, in support of the Oregon Governor’s Advisory Group on Global Warming

2. Why they did it: In 2004, Oregon’s Governor convened an Advisory Group on Global Warming. Seven technical subcommittees were formed to evaluate the feasibility and GHG reduction potential of policy and program options that the Advisory Group might recommend to the Governor. All subcommittees were required to project emissions through 2025 under a “business as usual” scenario, and then to project emissions through 2025 for each of the policy and program options. The difference between these two projections would be the emissions reductions associated with each policy or program.

This mandate posed some special challenges to the technical subcommittee on materials and waste.

  • First, many of the emissions potentially being reduced through recycling and prevention were not included in Oregon’s GHG inventory. Only including those emissions reductions that would be captured in Oregon’s inventory would be both challenging (e.g. some recyclables go to in-state markets, others are exported), and it would also undercount the full, global impacts of recycling and prevention in Oregon.
  • Second, the requirement that emissions (and emissions reductions) be identified for the year in which the emissions actually occur (or are reduced) required DEQ to add a temporal dimension to its model – simply using WARM would not work, since WARM sums all future emissions reductions (e.g. from forest carbon sequestration, avoided landfill gas production, etc.) into a single total. Click here for additional details regarding timing of emissions and emissions reductions.

3. What they did: DEQ started by building a model of waste generation for the “business as usual” scenario. This involved starting with baseline waste flow data for a base year (2002). Fortunately, Oregon has excellent data both for the quantities and composition of waste disposed, as well as the quantities of different types of wastes recovered via recycling and composting. Waste recovery and disposal categories didn’t exactly match but these data were combined into a profile of waste generation, recycling, composting, and disposal (by major disposal site) for 30 different categories of waste.

DEQ then created the “business as usual” forecast through 2025 by holding all recovery rates constant but increasing overall waste generation. Waste generation was forecast as a function of population and per-capita generation for each of the 30 categories. Rates of change in per-capita waste generation varied between waste categories (for example, per-capita generation of electronics waste was assumed to increase faster than most other categories). Using DEQ and EPA data, estimates were made of the rate of change in per-capita waste generation for the ten year period ending in the base year (2002). The rates of adjusted growth in per-capita waste generation (by category) were then related to the rate of growth in inflation-adjusted Oregon personal income during the same period. Looking forward, it was assumed that the relationships between personal income and per-capita waste generation would remain constant, in the "business as usual" scenario. Projections of population and inflation-adjusted personal income from the Oregon Department of Administrative Services were then used to project waste generation and flows through 2025.

The 30 materials categories were then related to material types in EPA’s WARM model. Emissions factors underlying the various stages of the life cycle were drawn from the supporting documentation. For all materials for which EPA provided a “source reduction” emissions factor, this factor was used to estimate “upstream”, or production-related, emissions. (Notably, source reduction emissions factors are missing at the time of this anaylsis for some major waste categories, including food and concrete.) Wastes diverted to recycling or composting were then given “credits” (emissions reductions) for all benefits except disposal avoidance. (Disposal avoidance was not credited as a negative emission, but since prevention, recycling and composting all reduce waste disposal, the related disposal-site emissions were reduced.) Remaining wastes being sent to disposal (via incineration or landfilling) were assigned emissions for incineration and landfilling. To the extent that both disposal methods offer GHG reductions/offsets (via energy production and also carbon storage in landfills), these reductions/offsets were also included.

The model includes waste disposed of in Oregon that is generated in other states (Oregon is a net importer of waste). Because Oregon can only directly influence the end-of-life emissions of these wastes, only their disposal-related emissions were included in the model.

Addressing the temporal nature of emissions and emissions reductions added to the complexity of the effort.

 

  • All wastes in the model are “generated”, by definition. DEQ used “manufacturing process energy”, “manufacturing transportation energy”, and “manufacturing process non-energy” emissions factors for these emissions, and assigned them all to the year in which the waste was generated. This is a reasonable approximation for wastes that are generated shortly after production (e.g. newspaper) but a very poor approximately for products that have long service lives (e.g. building materials). For wastes that undergo “waste prevention”, generation-related emissions are set to zero.
  • In addition, there is a forest carbon sequestration credit, consistent with WARM. This credit is allocated in equal portions to each of 15 years beginning with the year in which the waste prevention occurs.
  •  For wastes that undergo recycling, the credits in WARM for “manufacturing process energy”, “manufacturing transportation energy”, and “manufacturing process non-energy” are credited to the year in which the recycling activity occurs. Forest carbon sequestration credits are allocated in equal portions to each of 15 years beginning with the year in which the recycling activity occurs.
  • For wastes that undergo composting, emissions for transportation and equipment are assigned to the year in which the composting occurs. Emissions credits for soil carbon restoration and humus formation are allocated in equal portions to each of 15 years beginning with the year in which the composting activity occurs.
  • All emissions associated with combustion of waste (transportation/equipment, direct emissions, and energy offset credits) are assigned to the year in which the waste is combusted.
  • DEQ used a waste-in-place approach for landfill-related emissions. For emissions resulting from waste disposed of prior to 2002 ("legacy wastes"), DEQ used estimates of historic waste-in-place and EPA’s LandGem model to estimate methane generation by year (through 2025). Methane generation in any given year (e.g. 2005) from the ongoing decay of "legacy" wastes was then added to methane generation from waste disposed in intervening years (2002, 2003, 2004) for a total of methane generation by year by landfill. For waste placed in 2002 and subsequent years, DEQ used methane generation factors in WARM to estimate total lifetime generation, and then allocated these emissions using methane generation curves (for wet vs. dry landfills). Collection efficiencies and oxidation rates were used to project gas capture (for flaring and/or energy recovery) and gas emissions. Methane generation curves were also used to allocate landfill carbon storage credits over the years following disposal.

For additional details on the methodology, see pages 100 - 105 of:
http://oregon.gov/ENERGY/GBLWRM/docs/GWReport-FInal.pdf.

4. Results/outcomes/successes/failures/lessons learned:

The modeling approach used by DEQ illustrated that the “life cycle” emissions associated with waste generation in Oregon were significant, totaling approximately 10% of emissions in the State’s traditional inventory. However, it was also recognized that DEQ’s method significantly undercounted emissions. Major reasons include the following:

  • Only those materials that end up contributing to “solid waste generation” were counted. Most significantly, this excluded the majority of food purchases.
  • For materials for which “upstream” (production) related emissions factors are not included in WARM, emissions were treated as zero. These included high-volume, high-emissions materials such as food and concrete.
  • Waste generation was used as a proxy for consumption. This is a reasonable assumption for products with short service lives, but not for products with long service lives (construction materials).
  • The WARM emissions factors are derived from North American manufacturing data. The effect is to treat all foreign-made goods as having the same emissions impacts as goods made domestically.

Despite these limitations, the effort succeeded in demonstrating the moderate – and typically overlooked – impacts of materials use and waste generation. The model was also successful at evaluating the reductions in future GHG emissions resulting from hypothetical programs and policies. The evaluation confirmed the potential to reduce GHG emissions through a variety of approaches. Ultimately, the Governor’s Advisory Group on Global Warming forwarded ten recommendations related to materials. Three of them were classified as top priority: achieving the state’s existing waste generation and recycling goals (5.2 million MT CO2e reduction in 2025), and two measures specific to landfill gas (1.00 – 1.41 million MT CO2e reduction in 2025). On the first point, it should be noted that Oregon has both statutory recovery and statutory prevention goals; the modeling suggested that while recovery (recycling, composting) could contribute to further GHG reductions, the emissions reduction would be many, many times higher if the statutory prevention goal could be achieved. This reflects the fact that the “business as usual” (baseline) scenario was already fairly close to achieving the statewide recovery goal, while the difference between the “business as usual” scenario and statutory prevention goals was very wide.

The greatest challenge of this effort was the complexity of the model and the amount of time required to create it.

Because the model included emissions that were already captured in Oregon’s traditional inventory, DEQ did not try to integrate the modeled results into the traditional inventory. Rather, the results were held aside as a “side calculation”, and it was noted that the emissions and emissions reductions had some partial overlap with the conventional inventory.