Analysis of a real-time signal for greenhouse gas emissions of district heating consumption

Abstract: The district heating system (DHS) of Stockholm is one of the largest systems in the world with a total yearly production of 10TWh of heat and 2TWh of electricity (through combined heat and power plants). Large amounts of greenhouse gasses (GHG) are emitted to produce this heat and electricity. Given the goal of the City of Stockholm to reduce the amount of GHG emissions to 3 ton per capita in 2015 and to keep reducing emissions at a similar rate after 2015, it is important to identify the potentials for further reductions. Numerous studies have been done on how the DHS can become more sustainable by installing new generation units. However, also the consumers have an influence on the DHS. After all, it are the consumers who decide when and how much heat or electricity they use. Most former studies and environmental guidelines for the DHS in Stockholm focussed on the producer side. This thesis looks at the consumer perspective of the (heat of the) district heating system. A real-time signal giving the greenhouse gas emissions of individual households is developed and its potential and challenges are discussed. With this signal, households that want to minimise their environmental impact have a tool to decrease their environmental impact by changing their consumption. This can be a first step to transform the DHS to a smart district heating system. First, generic models to calculate the dynamic greenhouse gas intensity of the heat production of district heating and to calculate the greenhouse gas emissions related to the heat consumption of households are suggested. Then the feedback signal with those real-time household emissions is calculated for representative households in Stockholm based on data of Stockholm’s DHS and data about hot tap water consumption in Sweden. Results indicate that variations in household level greenhouse gas emissions mostly reflect changes in consumption but can also result from changes by the producer. Intraday variations are mostly caused by changes in hot tap water consumption, while variations on a timescale of a few days are caused by changes in heating consumption (changing weather) and changes made by the producer (to use different fuels). Then several scenarios are calculated, each scenario looking at the actions a consumer can take to shift or reduce his/her consumption (decrease hot tap water usage, lower indoor temperature etc.). The real-time household emissions are calculated again to see if the signal gives the needed incentives (is the household rewarded for its effort? Does it get further incentives?). It was found that a strong time-incentive (to decrease consumption when it saves most emissions) is missing if the average perspective is used to calculate the emission intensity of the heat production. Also, the results confirmed the finding that the feedback signal might not reflect changes in consumption. Finally, challenges for the signal are discussed. One of the major hurdles is the fact that household consumption of heat (heating and hot tap water) can often not be measured on a household level. Thus, it has to be estimated but it seems very difficult to get this estimation accurate enough to give correct feedback to households, especially about the emissions saved by their efforts to reduce/shift their consumption. Secondly, the time resolution should be chosen well to still get accurate results but not make the signal to data-intensive. Finally, the result is heavily dependent on the chosen methodology (average or marginal perspective? Do you account for the electrical side of the DHS? How about the distribution losses? Etc.).