Solid Waste Life-Cycle Modeling Wake County Solid Waste Management Division & NC State University Solid Waste Life-Cycle Modeling November 1, 2017

New Convenience Center Statement of Purpose The Solid Waste Division of Environmental Services protects the public health and safety of Wake County citizens by providing quality solid waste and recycling services that are efficient, cost effective and environmentally responsible. Landfill Gas Blower Landfill Partial Closure New Convenience Center Under Construction

Scope of Solid Waste Services Reduce Waste Increase recycling & reduce litter thru education & outreach Provide disposal location via SWLF and EWTS Minimize illegal dumping thru enforcement & providing convenience centers for residential use Collect banned items from landfills at MMRFs for business & residential use Protect the environment by providing HHW facilities for residential use and monitoring and maintaining closed landfills Increase recycling & reduce litter thru education & outreach Provide disposal location via SWLF and EWTS Minimize illegal dumping thru enforcement & providing convenience centers for residential use Collect banned items from landfills at MMRFs for business & residential use Protect the environment by providing HHW facilities for residential use and monitoring and maintaining closed landfills

Ongoing & Upcoming Issues Population Growth Over 1 million residents More People=More Waste Landfill Life Potentially Shortened

Growth & Sustainability Current Board of Commissioners established Board goals during 2016 GS2.3 - Extend the life of the South Wake Landfill and development of SW Comprehensive Plan Better Recycling Use Technology & Innovation, including modeling of future waste scenarios and options Life of landfill (LOL…) study underway via SCS with coordination efforts with NCSU modeling SW Action Plan reflects ongoing & upcoming efforts

Wake County Solid Waste Action Plan

Life-Cycle Modeling Decision Wake County, via our SWLF Partnership, provides MSW disposal services for munis & commercial waste haulers Though many years of capacity remain for the SWLF (2040+), due to timing to consider new landfill or other waste disposal options, determined a need to start investigating now Prior studies (1995, 2003) looked at specific technology Decision to develop model that could be used to evaluate all aspects (cost, resource/energy consumption and environmental performance) that can be continuously updated to reflect changing assumptions

NCSU Partnership Established contract with NCSU Civil Engineering Department to expand development of SWOLF specifically for Wake County use Work started in 2015 with extensive data collection DEQ Annual Reports County Solid Waste Management Plan Meetings with munis and county staff

Problem Statement Evaluate strategies to cost-effectively improve the sustainability of Wake County’s municipal solid waste management while considering changing population, waste generation and composition, landfill life, energy and material recovery, and environmental emissions and impacts

Research Approach Use a tool developed at NC State Solid Waste Optimization Lifecycle Framework (SWOLF) SWOLF estimates the full system costs and emissions associated with waste management processes collection through final disposal considers benefits from recycling and energy recovery Modeled Wake County’s solid waste system based on available data and reports waste generation and composition existing and potential future waste management facilities Costs aren’t just for the county. Includes collection, MRF operations, composting, etc.

SWOLF – Solid Waste Optimization Lifecycle Framework Evaluate system performance (i.e., economical, environmental) while accounting for changes to waste composition and generation, SWM policy, the U.S. energy system, and potential future GHG mitigation policies LCA Model Impact Assessment Model (e.g., Global Warming, Smog Formation) GHG Policy Funding from NSF and EREF Grants. I was funded as an EREF Scholar. SWM Process Models Optimizable Integrated SWM System Model Cost Emissions Energy Use Impacts Energy System SWM Policy Waste Generation and Composition

Benefits of Optimization Modeling (1/2) How can net present cost be minimized over time? While meeting diversion or greenhouse gas constraints Considering existing infrastructure How can environmental benefits be maximized? Minimize greenhouse gas emissions Minimize fossil energy use Maximize landfill diversion Impose a budget constraints

Benefits of Optimization Modeling (2/2) What are the mitigation costs ($/MTCO2E avoided) or trade- offs associated with adopting a specific technology or policy? WTE combustion, composting, AD, etc. Landfill organics bans, diversion targets, combustion How do changes to the energy system affect these decisions? Can our system robustly adapt to changes to the energy system, policy, waste composition, and waste generation? if gasification is mentioned, perhaps mention it is not embedded in SWOLF but we have done some LCA work on it. Gasification will

Representing Wake County in SWOLF Existing facility options Composting facilities (COMP) Landfill (LF) Single-stream material recovery facility (SSMRF) Transfer station (TS) Residential waste 10 single family sectors 2 multi-family sectors Convenience (drop off) Centers Collection or drop-off Recyclables Organics (yard, food waste) Residual waste Future facility options Anaerobic digestion (AD) Thermal waste-to-energy (WTE) Mixed waste material recovery facility (MWMRF) The SWOLF model represents a SWM system from waste generation, through collection or drop-off, and then treatment and final disposal. This figure illustrates the processes that are included in SWOLF. Existing facilities are highlighted in blue, and potential new facilities are highlighted in green. SWOLF is built on individual models that represent each of the processes on the screen, including collection and each existing or future facility. We can modify each of these to better reflect the system we are evaluating. Waste generation sectors Incorporated single-family residential Modeled as 10 single-family sectors (for these results; currently use 12) Multi-family residential Modeled as 2 multi-family sectors Unincorporated residential Modeled as 1 convenience center sector Not modeled: Unincorporated, private collection (little data) Commercial (little data, county does not control) Private waste collection Commercial waste Sector specific

Representing Wake County in SWOLF Waste generation and composition Changes to one process can have system-wide effects For example, adding residential food waste collection will affect the collection system, landfill life, greenhouse gas emissions, and energy recovery, and composting facilities Potential Facilities Existing Facilities

Model Objective: Least Cost Scenarios Developed several scenarios to explore how new processes could be added to the existing solid waste system. Scenarios Description Model Objective: Least Cost Current practice (Base) Separate collection of recyclables going to single-stream MRF and yard waste going to compost (2) Current + food waste collection (+FW) As in case 1 plus food waste co-collected with yard waste (3) Current + food waste collection + AD enabled (+FWAD) As in case 2 plus AD enabled (4) Case 3 + MWMRF As in case 1 plus mixed waste MRF (MWMRF) and AD enabled (5) Case 3 + WTE As in case 1 plus WTE combustion and AD enabled (6) Case 3 + WTE + MWMRF (ALL) As in case 1 plus WTE, MWMRF, and AD enabled Used food waste collection in all scenarios since it is a marginal difference (no difference in collection cost since it averages out to be the same). Small difference in treatment costs/diversion/emissions.

Adding food waste collection to current system Adding food waste to yard waste collection has a small impact on the average cost of collection ($/Mg) Discuss odors or specifics of increasing costs. The composting type used in the model was windrow - point out windrows without odor control works for only a limited amount of food waste and will need odor control and active aeration as food waste composting increases.  If the composting system is more advanced, mention what it is Discuss that food waste doesn’t ultimately take up much landfill space, so while the diversion increase is observed, the impact on landfill life would be small. Discuss that collection dominates costs and emissions. With the linear model, as waste quantity goes up for a waste stream (e.g., more recyclables collection), the per ton cost goes down, and vice versa (e.g., residual quantity goes down and cost per ton for residual collection goes up). This analysis assumes that the total collection cost is the same, which is a plausible scenario. More detailed analysis of possible collection changes could be performed that consider changes in collection frequency or other operational changes. The revised analysis performed for the manuscript included Non-vegetable food waste, which increased the diversion amount. It was still assumed that the separation efficiency was 50% Base +FW +FWAD +FWAD Min GHG

Adding food waste collection to current system Adding food waste to yard waste collection has a small impact on the average cost of collection ($/Mg) Same; AD not used if minimizing cost Discuss odors or specifics of increasing costs. The composting type used in the model was windrow - point out windrows without odor control works for only a limited amount of food waste and will need odor control and active aeration as food waste composting increases.  If the composting system is more advanced, mention what it is Discuss that food waste doesn’t ultimately take up much landfill space, so while the diversion increase is observed, the impact on landfill life would be small. Discuss that collection dominates costs and emissions. With the linear model, as waste quantity goes up for a waste stream (e.g., more recyclables collection), the per ton cost goes down, and vice versa (e.g., residual quantity goes down and cost per ton for residual collection goes up). This analysis assumes that the total collection cost is the same, which is a plausible scenario. More detailed analysis of possible collection changes could be performed that consider changes in collection frequency or other operational changes. The revised analysis performed for the manuscript included Non-vegetable food waste, which increased the diversion amount. It was still assumed that the separation efficiency was 50% Base +FW +FWAD +FWAD Min GHG

Adding food waste collection to current system Adding food waste to yard waste collection has a small impact on the average cost of collection ($/Mg) GHGs Decrease; AD Utilization Discuss odors or specifics of increasing costs. The composting type used in the model was windrow - point out windrows without odor control works for only a limited amount of food waste and will need odor control and active aeration as food waste composting increases.  If the composting system is more advanced, mention what it is Discuss that food waste doesn’t ultimately take up much landfill space, so while the diversion increase is observed, the impact on landfill life would be small. Discuss that collection dominates costs and emissions. With the linear model, as waste quantity goes up for a waste stream (e.g., more recyclables collection), the per ton cost goes down, and vice versa (e.g., residual quantity goes down and cost per ton for residual collection goes up). This analysis assumes that the total collection cost is the same, which is a plausible scenario. More detailed analysis of possible collection changes could be performed that consider changes in collection frequency or other operational changes. The revised analysis performed for the manuscript included Non-vegetable food waste, which increased the diversion amount. It was still assumed that the separation efficiency was 50% Base +FW +FWAD +FWAD Min GHG

Adding food waste collection to current system Adding food waste to yard waste collection has a small impact on the average cost of collection ($/Mg) Cost Increases Discuss odors or specifics of increasing costs. The composting type used in the model was windrow - point out windrows without odor control works for only a limited amount of food waste and will need odor control and active aeration as food waste composting increases.  If the composting system is more advanced, mention what it is Discuss that food waste doesn’t ultimately take up much landfill space, so while the diversion increase is observed, the impact on landfill life would be small. Discuss that collection dominates costs and emissions. With the linear model, as waste quantity goes up for a waste stream (e.g., more recyclables collection), the per ton cost goes down, and vice versa (e.g., residual quantity goes down and cost per ton for residual collection goes up). This analysis assumes that the total collection cost is the same, which is a plausible scenario. More detailed analysis of possible collection changes could be performed that consider changes in collection frequency or other operational changes. The revised analysis performed for the manuscript included Non-vegetable food waste, which increased the diversion amount. It was still assumed that the separation efficiency was 50% Base +FW +FWAD +FWAD Min GHG

Cases 1-3: Base case and adding food waste collection to current system Adding food waste to yard waste collection has a small impact on the average cost of collection ($/Mg) GHG Mitigation cost for adding food waste collection is 550 $/MTCO2e Minimizing GHG with food waste collection and AD increases cost by $3.1M with a mitigation cost of $920 MTCO2e Discuss odors or specifics of increasing costs. The composting type used in the model was windrow - point out windrows without odor control works for only a limited amount of food waste and will need odor control and active aeration as food waste composting increases.  If the composting system is more advanced, mention what it is Discuss that food waste doesn’t ultimately take up much landfill space, so while the diversion increase is observed, the impact on landfill life would be small. Discuss that collection dominates costs and emissions. With the linear model, as waste quantity goes up for a waste stream (e.g., more recyclables collection), the per ton cost goes down, and vice versa (e.g., residual quantity goes down and cost per ton for residual collection goes up). This analysis assumes that the total collection cost is the same, which is a plausible scenario. More detailed analysis of possible collection changes could be performed that consider changes in collection frequency or other operational changes. The revised analysis performed for the manuscript included Non-vegetable food waste, which increased the diversion amount. It was still assumed that the separation efficiency was 50% Base +FW +FWAD +FWAD Min GHG

Case 4: + Mixed Waste MRF (+MWMRF) Current system + food waste collection with AD enabled (optional) + mixed waste MRF Separate collection of recyclables and yard/food waste required (3 separate collections) Set increasing diversion targets Lowest target = diversion in min-cost solution Highest target = diversion in max-diversion solution

Case 4: + Mixed Waste MRF (+MWMRF) Mixed waste MRF use starts with sectors closet to MRF; furthest from LF Adding FW collection provided CO2 mitigation at a cost of $0.55/kg (going from Case 1 to Case 4, least cost). You can see in the least cost solution, neither MWMRF nor AD was used, so this is equivalent to the min cost for cases 2 and 3. As the diversion target was increased, some single family sectors begin using MWMRF over LF to meet the diversion goal. In terms of CO2 mitigation, the mitigation cost starts at $0.09/kg and reduces to about 5 cents per kg at 38% diversion. At max diversion, the mitigation cost gost up slightly to 5.1 cents/kg CO2. As the diversion target was increased, some single family sectors begin using MWMRF over LF to meet the diversion goal. This starts with sectors that are closer to the MRF than the landfill (starting with those with the biggest difference, then going towards those that are almost the same distance). SF 3,4,8,10 then MF starts using partial MWMRF, then CC starts using MWMRF, and last to switch are those sectors that are farther away from the MWMRF than they are to the landfill. Collection cost drives the transition from LF to MWMRF Base Percent Diverted

Case 4: + Mixed Waste MRF (+MWMRF) Diversion, MWMRF Use, and Cost Increase GHG Emissions Decrease Mitigation cost: 50 to 90 $/MTCO2e Adding FW collection provided CO2 mitigation at a cost of $0.55/kg (going from Case 1 to Case 4, least cost). You can see in the least cost solution, neither MWMRF nor AD was used, so this is equivalent to the min cost for cases 2 and 3. As the diversion target was increased, some single family sectors begin using MWMRF over LF to meet the diversion goal. In terms of CO2 mitigation, the mitigation cost starts at $0.09/kg and reduces to about 5 cents per kg at 38% diversion. At max diversion, the mitigation cost goest up slightly to 5.1 cents/kg CO2. As the diversion target was increased, some single family sectors begin using MWMRF over LF to meet the diversion goal. This starts with sectors that are closer to the MRF than the landfill (starting with those with the biggest difference, then going towards those that are almost the same distance). SF 3,4,8,10 then MF starts using partial MWMRF, then CC starts using MWMRF, and last to switch are those sectors that are farther away from the MWMRF than they are to the landfill. Collection cost drives the transition from LF to MWMRF Base Percent Diverted

Case 5: + WTE Combustion (+WTE) Current system + food waste collection with AD enable (optional) + WTE Separate collection required Set increasing diversion targets Lowest target = diversion in min-cost solution Highest target = diversion in max-diversion solution

Case 5: + WTE Combustion (+WTE) WTE (located at LF) starts with selected sectors based on composition Adding FW collection provided CO2 mitigation at a cost of $0.55/kg (going from Case 1 to Case 5, least cost). You can see in the least cost solution, neither WTE nor AD was used, so this is equivalent to the min cost for cases 2 and 3. (and case 4) As the diversion target was increased, some single family sectors begin using WTE over LF to meet the diversion goal. The mitigation cost starts at $0.176/kg and reduces to about 15.7 cents per kg at 70% diversion. At max diversion, the mitigation cost goes up to 17 cents/kg CO2. The WTE facility was assumed to be at the SWLF location, so the distance from each sector to each site is the same. Some SF sectors are transferred first. By 60% diversion, all MF and CC waste is at WTE, and lastly some SF sectors are brought to WTE. MWMRF transition appeared driven by collection costs, but in this case collection costs should be equal, so must be driven by composition differences in the waste that select the waste for which the additional costs from WTE treatment over LF treatment are lowest. This is more complex than just recycling value, it likely is made up of how much LFG is eliminated by going to WTE vs the heat value of that material, and the revenue from recovered metals. Base Percent Diverted

Case 5: + WTE Combustion (+WTE) Diversion, WTE Use, and Cost Increase GHG Emissions Decrease Mitigation cost: 160 to 180 $/MTCO2e Adding FW collection provided CO2 mitigation at a cost of $0.55/kg (going from Case 1 to Case 5, least cost). You can see in the least cost solution, neither WTE nor AD was used, so this is equivalent to the min cost for cases 2 and 3. (and case 4) As the diversion target was increased, some single family sectors begin using WTE over LF to meet the diversion goal. The mitigation cost starts at $0.176/kg and reduces to about 15.7 cents per kg at 70% diversion. At max diversion, the mitigation cost goes up to 17 cents/kg CO2. The WTE facility was assumed to be at the SWLF location, so the distance from each sector to each site is the same. Some SF sectors are transferred first. By 60% diversion, all MF and CC waste is at WTE, and lastly some SF sectors are brought to WTE. MWMRF transition appeared driven by collection costs, but in this case collection costs should be equal, so must be driven by composition differences in the waste that select the waste for which the additional costs from WTE treatment over LF treatment are lowest. This is more complex than just recycling value, it likely is made up of how much LFG is eliminated by going to WTE vs the heat value of that material, and the revenue from recovered metals. Base Percent Diverted

Case 6: + MWMRF + WTE (ALL) Current system + food waste collection with AD enabled (optional) + mixed waste MRF + WTE Separate collection of recyclables and yard/food waste required (3 separate collections) Set increasing diversion targets Lowest target = diversion in min-cost solution Highest target = diversion in max-diversion solution

Case 6: +MWMRF + WTE (ALL) MWMRF is only used to meet maximum diversion Adding FW collection provided CO2 mitigation at a cost of $0.55/kg (going from Case 1 to Case 6, least cost). You can see in the least cost solution, neither MWMRF, WTE nor AD was used, so this is equivalent to the min cost for cases 2 and 3. (and case 4 and 5) As the diversion target was increased, some single family sectors begin using a combination of MWMRF and WTE over LF to meet the diversion goal. The mitigation cost starts at $0.176/kg and reduces to about 15.7 cents per kg at 70% diversion. At max diversion, the mitigation cost goes down to 12.4 cents/kg CO2. The WTE facility was assumed to be at the SWLF location, so the distance from each sector to each site is the same. Some SF sectors are transferred first. By 60% diversion, all MF and CC waste is at WTE, and lastly some SF sectors are brought to WTE. MWMRF transition appeared driven by collection costs, but in this case collection costs should be equal, so must be driven by composition differences in the waste that select the waste for which the additional costs from WTE treatment over LF treatment are lowest. This is more complex than just recycling value, it likely is made up of how much LFG is eliminated by going to WTE vs the heat value of that material, and the revenue from recovered metals. Base Percent Diverted

Case 6: +MWMRF + WTE (ALL) Increasing WTE Use Adding FW collection provided CO2 mitigation at a cost of $0.55/kg (going from Case 1 to Case 6, least cost). You can see in the least cost solution, neither MWMRF, WTE nor AD was used, so this is equivalent to the min cost for cases 2 and 3. (and case 4 and 5) As the diversion target was increased, some single family sectors begin using a combination of MWMRF and WTE over LF to meet the diversion goal. The mitigation cost starts at $0.176/kg and reduces to about 15.7 cents per kg at 70% diversion. At max diversion, the mitigation cost goes down to 12.4 cents/kg CO2. The WTE facility was assumed to be at the SWLF location, so the distance from each sector to each site is the same. Some SF sectors are transferred first. By 60% diversion, all MF and CC waste is at WTE, and lastly some SF sectors are brought to WTE. MWMRF transition appeared driven by collection costs, but in this case collection costs should be equal, so must be driven by composition differences in the waste that select the waste for which the additional costs from WTE treatment over LF treatment are lowest. This is more complex than just recycling value, it likely is made up of how much LFG is eliminated by going to WTE vs the heat value of that material, and the revenue from recovered metals. Base Percent Diverted

Case 6: +MWMRF + WTE (ALL) Mitigation cost is 120 to 180 $/MTCO2e Adding FW collection provided CO2 mitigation at a cost of $0.55/kg (going from Case 1 to Case 6, least cost). You can see in the least cost solution, neither MWMRF, WTE nor AD was used, so this is equivalent to the min cost for cases 2 and 3. (and case 4 and 5) As the diversion target was increased, some single family sectors begin using a combination of MWMRF and WTE over LF to meet the diversion goal. The mitigation cost starts at $0.176/kg and reduces to about 15.7 cents per kg at 70% diversion. At max diversion, the mitigation cost goes down to 12.4 cents/kg CO2. The WTE facility was assumed to be at the SWLF location, so the distance from each sector to each site is the same. Some SF sectors are transferred first. By 60% diversion, all MF and CC waste is at WTE, and lastly some SF sectors are brought to WTE. MWMRF transition appeared driven by collection costs, but in this case collection costs should be equal, so must be driven by composition differences in the waste that select the waste for which the additional costs from WTE treatment over LF treatment are lowest. This is more complex than just recycling value, it likely is made up of how much LFG is eliminated by going to WTE vs the heat value of that material, and the revenue from recovered metals. Base Percent Diverted

Observations and Summary SWOLF is effective at quantifying the cost and environmental impacts of new technology implementation and associated tradeoffs between cost, diversion and GHGs Continuous engagement with the county was very important in the development of useful results The county does not control municipal collection or commercial waste Dialog with municipalities is essential for long-term facility commitments Impact of commercial waste on facility sizing and life is required Food waste is dense, has a high moisture content, and degrades relatively fully, so effect on landfill life is further reduced from the model results

Observations and Summary Collecting residential food waste with yard waste is predicted to decrease landfill greenhouse gas (GHG) emissions by 12%, but has a modest effect on diversion rate and landfill life. Increasing residential recycling participation will decrease GHG emissions and increase landfill diversion. Extension of landfill life is limited by Current waste composition (what can be diverted) Participation in recycling efforts Increasing diversion does not necessarily decrease GHG emissions Max diversion with Waste-to-Energy combustion is 81% Min GHG emissions occur with 77% diversion Food waste is dense, has a high moisture content, and degrades relatively fully, so effect on landfill life is further reduced from the model results

Next Steps Complete Life-Cycle Study/report Coordinate with Life of Landfill study Update on a regular basis to reflect system changes and monitor impacts – changes such as: Energy pricing System hauling/pick up changes Waste generation and composition

Questions? John Roberson Wake County Solid Waste Director john.roberson@wakegov.com Morton Barlaz North Carolina State University barlaz@ncsu.edu