Preliminary Greenhouse Gas (GHG) Emissions Analysis of Four Gates Sanitation Systems Emissions during Steady-State Operation By John T. Trimmer, Diana M. Byrne, Hannah A.C. Lohman, & Jeremy S. Guest University of Illinois at Urbana-Champaign Prepared for: Duke Internal Subaward #: TO 283-1325 Brian Stoner, Ph.D. Center for WaSH-AID [email protected] Primary Author Contact Information: Date Submitted: 12/20/2018 Jeremy S. Guest, Ph.D. Assistant Professor Department of Civil and Environmental Engineering University of Illinois at Urbana-Champaign 3221 Newmark Civil Engineering Laboratory, MC-250 205 North Mathews Avenue Urbana, IL 61801-2352 [email protected] BMGF OPP: OPP1173370 Project #: 13C ________________________________________________________________________________________________________ This work is, in part, supported by a grant, OPP1173370, from the Bill & Melinda Gates Foundation through Duke University’s Center for WaSH-AID. All opinions, findings, and conclusions or recommendations expressed in this work are those of the author(s) and do not necessarily reflect the views of the Foundation, Duke, or the Center. Statement of Confidentiality This report and supporting materials contain confidential and proprietary information. These materials may be printed or photocopied for use in evaluating the project, but are not to be shared with other parties without permission. Jeremy S. Guest, Ph.D. 2 Assistant Professor Department of Civil and Environmental Engineering University of Illinois at Urbana-Champaign 3221 Newmark Civil Engineering Laboratory, MC-250 205 North Mathews Avenue Urbana, IL 61801-2352 Preliminary Greenhouse Gas (GHG) Emissions Analysis of Four Gates Sanitation Systems Emissions during Steady-State Operation John T. Trimmer, Diana M. Byrne, Hannah A.C. Lohman, Jeremy S. Guest Executive Summary This analysis estimates greenhouse gas (GHG) emissions associated with steady-state operation of four technologies in the Bill & Melinda Gates Foundation’s sanitation portfolio: the HTClean system (Helbling), the Empower Sanitation Platform (Duke Centre for WaSH-AID), the Electrochemical Reinvented Toilet (Eco-San), and the Janicki Omni-Processor (Sedron Technologies). Specifically, we estimated emissions from (i) the degradation of bodily waste during containment, treatment, and recovery; (ii) electricity and materials consumed during operation; and (iii) transportation of waste to its treatment and/or disposal site. We also estimated GHG offsets from recovery of fertilizer nutrients (nitrogen, phosphorus, potassium). This steady- state analysis does not include the emissions associated with the construction, maintenance, and end-of-life of these technologies. All estimates are expressed as equivalent kilograms of carbon -1 -1 dioxide per year, normalized to the estimated population served (i.e., kg CO2 eq·cap ·year ). With the exception of the HTClean system, direct emissions from the current portfolio technologies have the potential to compare favorably against pit latrines. Our results suggest that electricity demand tends to drive emissions trends across three of the four systems (excluding the Omni-Processor). HTClean is associated with the highest total GHG emissions per person -1 -1 (780-1,800 kg CO2 eq·cap ·yr ), with 90% of emissions coming from its large electricity demand. As the Omni-Processor is reported to require no outside electricity, it is associated with the lowest -1 -1 emissions (33-64 kg CO2 eq·cap ·yr for the full system including latrine containment and passive pretreatment). Latrine containment and pretreatment contribute most of the emissions related to -1 -1 the Omni-Processor system. The Empower (100-180 kg CO2 eq·cap ·yr ) and Eco-San (250- -1 -1 470 kg CO2 eq·cap ·yr ) systems produce intermediate levels of emissions (compared with the HTClean and Omni-Processor systems), with the Eco-San total being larger due to its higher electricity demand. The Empower system offers two alternatives for final processing of dried solids (combustion or land application), but total emissions are similar for both options (as land-applied solids may continue to degrade). The Empower estimates are similar to preliminary values associated with degradation of bodily waste in pit latrines (as calculated by the University of Leeds CACTUS team in ongoing work; final values may differ slightly). In all full systems, recovered nutrients contribute offsets that are relatively small compared with total system emissions. Broadly, these preliminary results are associated with large uncertainties as the analysis required numerous assumptions. Results tend to be most sensitive to the emissions factor for electricity -1 (i.e., kg CO2 eq·kWh ) and multiple parameters related to direct emissions from bodily waste (e.g., carbon and nitrogen excretion, emissions during combustion). Electricity emissions will vary depending on the local source of electricity (e.g., coal, hydroelectric, solar), resulting in significant variability across implementation locations. This variability can be reduced if a power source is built into the system itself, and we show that integrating renewable energy sources (e.g., photovoltaics) may provide the greatest opportunity for reductions in total GHG emissions. Direct emissions vary based on storage and treatment conditions of bodily waste and will also depend upon implementation location, as local diets will affect excretion of carbon and nitrogen into sanitation systems. Moving forward, we hope to expand these analyses in collaboration with the design and modeling teams and develop a full life cycle assessment (LCA) of each system. 12/20/2018 BMGF OPP1173370, Project 13C (Duke TO 283-1325) Guest Introduction This report presents the preliminary findings from an analysis of greenhouse gas (GHG) emissions associated with four technologies in the Bill & Melinda Gates Foundation’s sanitation portfolio. The four technologies include the HTClean system (Helbling), the Empower Sanitation Platform (Duke Centre for WaSH-AID), the Electrochemical Reinvented Toilet (Eco-San), and the Janicki Omni-Processor S250 (Sedron Technologies), providing an array of systems ranging from a decentralized, single-household toilet up to a centralized, community-scale treatment facility. For this preliminary report, we present estimates of steady-state, operation phase emissions for each system. Three general categories of emissions are defined: • Direct emissions from degradation of bodily waste (biogenic methane and N2O given off during containment, treatment, and recovery; biogenic carbon dioxide is not included); • Technology operation (emissions associated with electricity and consumables needed for system functioning; replacement parts are not included); and • Transportation (truck conveyance of latrine sludge to centralized treatment facilities or recovered products to locations for land application). We also estimated offsets associated with nutrients (nitrogen, phosphorus, potassium) recovered in the products from each system. These products could replace conventional fertilizers and the emissions associated with their production. All emissions and offset estimates were converted to equivalent kilograms of carbon dioxide per year and were normalized according to the population -1 -1 served by each system (kg CO2 eq·cap ·yr ). In the future, we hope these analyses can be expanded to further inform decision-making through a full life cycle assessment of each system. Results and Discussion Estimates of greenhouse gas emissions during steady-state operation. To account for the uncertainties associated with our analysis, we present expected emissions estimates (calculated using single likely values of all assumed parameters) along with the results of an uncertainty analysis (in which we varied parameter values across 10,000 simulations; see Methods for more detail). All reported ranges represent the 5th and 95th percentile output values from this analysis. Generally, electricity demand tends to drive comparative performance concerning GHG emissions per person per year (Figure 1). The HTClean system is associated with the highest -1 -1 total emissions (780-1,800 kg CO2 eq·cap ·yr ), and 90% of emissions come from its large electricity demand. The Helbling Technik team (via Dr. Christian Seiler) estimates this system currently requires an average of 650 watts of continuous power. The average power draw of the next generation model may drop below 400 watts, representing a potential 40% reduction in energy-related emissions, but its total emissions would still be much larger than those of the other three systems. In the HTClean system, direct emissions during the high temperature/pressure treatment process are particularly uncertain, due to a lack of experimental data (we used estimates related to dehydration-stage pyrolysis1 at <200°C for methane and gaseous losses of 2 protein in hydrothermal liquefaction for N2O, placing large uncertainty ranges of ±100% around these values). However, the dominance of electricity emissions in this system reduces the importance of these tentative direct emissions assumptions. In contrast, the Omni-Processor is reported to require no outside electricity and is associated with the lowest emissions. We investigated three cases for the Omni-Processor: (i) emissions from the processor itself; (ii) emissions from the processor and a passive pretreatment approach that dewaters input sludge using unplanted drying beds and treats the resulting liquid effluent using anaerobic