As a result, CH4 production made up only 5% of net GHG emissions at Compost PAP and 48% of those from Compost CH. This finding suggests that aeration during the thermophilic phase of composting is critical in minimizing the GHG footprint of EcoSan systems. Waste treatment ponds produced anaerobic conditions that generate high levels of CH4 and very little CO2. The GHG contribution of waste stabilization ponds can be mitigated by the use of CH4 gas capture and electricity generation. Anaerobic digestion coupled to CH4 capture has been used to treat livestock manure and in some wastewater treatment plants for decades. However, many waste stabilization ponds throughout the world, including those sampled in this study, do not capture and reuse the CH4 generated during waste treatment. Market barriers – including the initial financial investment costs CH4 capture technology and electricity generation facilities – regulatory challenges, and lack of access to technology severely limit its widespread adoption in regions of the world that currently lack basic sanitation needs. Further, the efficiency of pathogen removal in waste stabilization ponds is highly variable , thereby limiting the effectiveness of waste stabilization ponds in regions of the world with limited technological and capital resources. Nitrous oxide is produced during the microbial-mediated processes of nitrification and denitrification, and can be produced in conditions with high to low levels of oxygen. Nitrification, the conversion of ammonium to nitrate through microbial oxidation, requires a source of ammonium and oxygen. During nitrification, N2O can form by the nitrate reductase enzyme in anaerobic conditions. Denitrification, the reduction of nitrate to dinitrogen through a series of intermediates, requires a source of nitrate, organic carbon, and limited oxygen.
Nitrous oxide can form as a result of incomplete denitrification to N2. Human waste contains organic carbon and a range of organic and inorganic forms of nitrogen. Therefore,vertical growing towers the oxygen conditions of a particular waste treatment pathway are a major control on N2O fluxes. In the anaerobic waste stabilization ponds, N2O was undetectable. In municipal wastewater treatment plants, measurements of N2O vary widely and can be mitigated by technologies that remove total nitrogen. Grass fields where waste was illegally disposed exhibited high and spatially variable N2O and CH4 fluxes. We observed a trade off between N2O and CH4 across sanitation pathways. Whereas waste stabilization ponds produced high levels of CH4 and no N2O, both EcoSan systems tended to have high fluxes of N2O. Nitrous oxide in compost piles can be produced by both nitrification and denitrification processes present along oxygen, moisture and C:N gradients within the pile. Reducing occurrences of anaerobic microsites could further limit N2O production from EcoSan compost, however, N2O production could still results from nitrification conditions. Despite this pollution swapping and taking into account the greater global warming potential of N2O, the largest contributor to GHG emissions from these systems is still CH4. Therefore, without systems in place to capture and oxidize CH4, the aerobic EcoSan system is a favorable system relative to waste stabilization ponds and illegal disposal on grass fields with respect to its impact on the climate.The management of aerobic biogeochemical conditions in compost piles plays a key role in minimizing CH4 and N2O losses. We observed large differences in CH4 emissions, and consequently in overall GHG emissions, across the two EcoSan systems in our study, implying opportunities for improved management. We tested this explicitly with a targeted comparison of GHG emissions above two piles, one with a permeable soil lining and one with an impermeable cement lining, at the Compost CH site and with a second comparison of GHG emissions before and after turning pile material. We found that CH4 emissions from the cement lined pile were approximately four times greater than from the soil lined pile, despite no significant temperature or CO2 emission differences.
This is evidence that higher CH4 emissions were driven by a larger methanogenic fraction , expressed as the amount of CH4 emitted per unit CO2, in the lined pile, indicating a greater prevalence of anaerobic conditions due to higher pile moisture. The cement-lined pile in the paired-pile experiment had no drainage mechanism and therefore likely represents a high endmember for wet pile conditions and high CH4 emissions. Notably, the standard design for cement-lined piles at the Compost CH4 site includes a lateral overflow PVC pipe, providing passive drainage, while at Compost PAP a soil lining is used without a PVC drain, and in both cases the CH4 emissions observed were much lower. The very high CH4 emissions from the undrained pile therefore likely reflect a very high moisture end-member for thermophilic composting. For future EcoSan implementations there are important trade offs to consider in pile design. The advantages of a PVC drain and associated storage tank are that potentially pathogenic liquid is contained, can be recycled to maintain optimal pile moisture levels under drier conditions and, if sanitized, the nutrient content of the leachate can be recycled. In contrast, a soil floor costs less, but it is important to consider, and monitor for, the potential leaching of pathogens, nutrients that can causes algalblooms, and trace metals that could contaminate drinking water when using a permeable floor. Future studies should further explore the quantity, composition, and timing of pile leaching, and assess the efficacy of soil as a filter to avoid contamination of groundwater alongside lowering GHG emissions. Though use of a permeable soil floor and/or PVC overflow drain showed potential to reduce EcoSan composting GHG emissions, the effects of turning the pile e even once e were even greater. Emissions of CH4 dropped two orders of magnitude, approaching zero, within one day after turning and stayed comparably low through the third day. Piles in the EcoSan second stage are turned every 7e10 days , therefore it is likely that CH4 emissions remain low throughout this entire phase, as originally evidenced by the >3-month time points in the initial measurements at Compost CH and Compost PAP. From these results, it may appear to be beneficial from a climate forcing perspective to reduce the time spent in the first static phase, however this must be balanced by the need to safely manage the pathogen burden at this early treatment stage, especially if piles are turned using manual labor.
Turning must only begin when pathogen abundance in the material has been reduced to a safe level, thus safeguarding the health of employees and local environment. Furthermore, though not observed in this study, past work has also shown that pile turning can increase N losses. Significant spikes in ammonia and N2O emissions follow mechanical turning of composting manure. It is therefore possible that within EcoSan composting there may be a trade-off between N2O and CH4 emissions between the initial static and later turned stages,container vertical farming similar to our observations across different sanitation pathways. Our gridded sampling scheme also allowed us to test the hypothesis that aeration drives CH4 emissions within piles. The results confirmed the utility of our model-based sampling design, with mean CH4 emissions four to five times higher from pile centers than pile corners or edges, regardless of the general drainage characteristics of the pile. An alternative to early turning may be the use of additional engineering to further aerate the middle of large piles where, even under well-drained pile conditions, we observed steep increases in CH4 emissions. One solution may be use of perforated PVC pipes for passive aeration of the pile at relatively low cost. Thermophilic composting is most effective under aerobic conditions. Understanding how management can best support aerobic conditions provides a win-win opportunity to increase the operational efficiency composting for treating waste while reducing the associated GHG emissions. The preliminary comparisons in this study captured significant effects of pile lining permeability and pile turning on GHG emissions during thermophilic composting, and helped us interpret the longer-term dynamics of GHG emissions during composting. Although our targeted measurements identify two of the management controls of GHG differences , robust estimates of emission factors for EcoSan composting requires a more comprehensive assessment of GHG dynamics, considering different management options, and with more extensive sampling throughout the composting operational stages. In sum these results support the potential for EcoSan composting to further reduce CH4 and overall GHG emissions associated with waste containment and treatment if piles are carefully designed and effectively managed to support aerobic metabolism.Strong consistency guarantees simplify programming complex, asynchronous distributed systems by increasing the number of assumptions a programmer can make about how a system will behave.
For years, system designers focused on how to provide the strongest possible guarantees on top of unreliable and even malicious systems. The rise of the Internet and cloud-scale computing, however, shifted the focus of system designers towards scalability. In a rush to meet the needs of cloud-scale workloads, system designers realized if they weakened the consistency guarantees they provided, they could greatly increase the scalability of their systems. As a result, designers simplified the guarantees provided by their systems and weaker consistency models such as eventual consistency emerged, greatly increasing the burden on developers. This movement towards weaker consistency and reduced features is known as NoSQL. NoSQL achieves scalability by partitioning or sharding data, spreading the load across multiple nodes. In order to maintain scalability, NoSQL designers ensured requests were not required to cross multiple partitions. As a result, they dropped traditional database features such as transactions in order to maintain scalability. While this worked for some applications, many of the developers with applications which needed this functionality were forced to choose between a database with all the functionality they needed, or to adapt their applications to the new world of the relaxed guarantees provided by NoSQL. Programmers found ways around the restrictions of weaker consistency by retrofitting transaction protocols on top of NoSQL, or by finding the minimum guarantees required by their application. Chapter 2 explores this pendulum away from and back towards consistency. This dissertation explores Corfu, a platform for scalable consistency. Corfu answers the question: “If we were to build a distributed system from scratch, taking into consideration both the desire for consistency and the need for scalability, what would it look like?”. The answer lies in the Corfu distributed log. Chapter 3 introduces the Corfu distributed log. Corfu achieves strong consistency by presenting the abstraction of a log – clients may read from anywhere in the log but they may only append to the end of the log. The ordering of updates on the log are decided by a high throughput sequencer, which we show can handle nearly a million requests per second. The log is scalable as every update to the log is replicated independently, and every append merely needs to acquire a token before beginning replication. This means that we can scale the log by merely adding additional replicas, and our only limit is the rate of requests the sequencer can handle. While Chapter 3 describes how to build a single distributed log, multiple applications may wish to share the same log. By sharing the same log, updates across multiple applications can ordered with respect to one another, which form the basic building block for advanced operations such as transactions. Chapter 4 details two designs for virtualizing the log: streaming, which divides the log into streams built using log entries which point to one another, and stream materialization, which virtualizes the log by radically changing how data is replicated in the shared log. Stream materialization greatly improves the performance of random reads, and allows applications to exploit locality by placing virtualized logs on a single replica. Efficiently virtualizing the log turns out to be important for implementing distributed objects in Corfu, a convenient and powerful abstraction for interacting with the Corfu distributed log introduced in Chapter 5. Rather than reading and appending entries to a log, distributed objects enable programmers to interact with in-memory objects which resemble traditional data structures such as maps, trees and linked lists. Under the covers, the Corfu runtime, a library which client applications link to, translates accesses and modifications to in-memory objects into operations on the Corfu distributed log. The Corfu runtime provides rich support for objects.