Cost-effective Planning of Decarbonized Power-Gas Infrastructure to Meet the Challenges of Heating Electrification

Abstract

Building heat electrification is central to economy-wide decarbonization efforts and directly affects energy infrastructure planning through increasing electricity demand and reducing the building sector's use of gas infrastructure that also serves the power sector. Here, we develop a modeling framework to quantify end-use demand for electricity and gas in the buildings sector under various electrification pathways and evaluate their impact on co-optimized bulk power-gas infrastructure investments and operations under deep decarbonization scenarios. Applying the framework to study the U.S. New England region in 2050 across 20 weather scenarios, we find high electrification of the residential sector can increase sectoral peak and total electricity demands by up to 56-158% and 41-59% respectively relative to business-as-usual projections. Employing demand-side measures like building envelope improvements under high electrification, however, can reduce the magnitude and weather sensitivity of peak load as well as induce overall efficiency gains, reducing the combined residential sector energy demand for power and gas by 28-30% relative to the present day. Notably, a combination of high electrification and envelope improvements yields the lowest bulk power-gas system cost outcomes. Accounting for \bt{midstream} methane emissions from gas supply chain increase the reliance on low-carbon fuels, which indirectly improves the cost-effectiveness of end-use electrification. Similarly, we find that demand flexibility programs can reduce the total system cost by up to 6.3%.

Figures (39)

• Figure 1: Overview of the modeling framework to evaluate the impact of residential heating electrification on energy infrastructure and operations planning outcomes. Residential demand is explicitly modeled in the study to consider the impact of various demand-side technological interventions. Non-residential power and gas demand is held constant across all scenarios evaluated as per the projections for high electrification scenarios available from another study NREL2021ElecReport.
• Figure 2: (a) Scenarios for electrification of residential building heating for 2050. "Winter-sized" homes have large whole-home heat pumps that are sized to provide heat through the winter and are used as the only source of heating and cooling. Smaller "Summer-sized" heat pumps are primarily sized to meet air conditioning needs in the summer, but also provide some heating in the winter, supplemented by a backup heating system. In our analysis, "Summer-sized" systems are almost all also hybrid systems because the existing heating systems in New England homes are typically gas or oil. We base our heat pump deployment scenarios on the Massachusetts Clean Energy and Climate Plan (CECP) for 2050, with our Medium Electrification scenarios corresponding to the CECP “Hybrid” scenario and the High Electrification scenarios corresponding to the CECP “High Electrification” scenario MA-climate-plan. In the MX and HX scenarios, we assume 70% of all homes given heat pumps also receive envelope improvements. See further details in SI \ref{['SIsec:deploymentscenarios']}. (b) Temperature profiles in Suffolk County, MA for the 20 weather years used in the analysis. Light blue lines in the background illustrate hourly temperature values for all 20 years overlaid on top of one another to illustrate the weather variations. Orange and bold red lines highlight weekly rolling averages of temperatures for two different years drawn from the dataset. A more detailed depiction of the 20 years of temperature data can be found in Fig. \ref{['SIfig:tempdists']}.
• Figure 3: (a) Annual electricity and gas demands in residential sector under different electrification scenarios. Each box plot contains demand data for the 20 weather years. HE, HX, ME, MX, and RF are all 2050 scenarios. The box edges correspond to interquartile range (IQR) and the whiskers indicate the most extreme points within the range Q1 - 1.5*IQR to Q3 + 1.5*IQR. (b) Summer and winter peak demands for the residential sector under different electrification scenarios. Each violin contains demand data simulated for 20 weather years. The width of plots increases where the density of data is higher, and its height represents sensitivity toward the weather variation. For New England, our "present-day" model and data workflow result in overestimation of residential gas consumption as compared to published outputs of ResStock runs by NREL ResstockComstock2022 as well as historical annual values available from EIA (shown in Figure \ref{['SIfig:histvspresent']}). As discussed in SI \ref{['SIsec:validation']}, the difference between our present-day model results and EIA data is likely attributable to a combination of a) differences in weather data used in our analysis (see Fig. \ref{['SIfig:ERA5vsNREL']}), b) the limited number of building archetypes we consider to maintain computational tractability, and c) error within the ResStock model. The analysis presented in \ref{['SIsec:validation']} suggests that the error in the present-day model does not as strongly impact the demand results from our future electrification scenarios.
• Figure 4: Average (statistical mean) residential peak load increase across the 20 weather years for scenario HX above present-day for 17 zones in New England. A summary of the zone geography is presented in SI \ref{['SIsec:Powernodes']}. Numerical results shown in Table \ref{['SItab:spatialdemand']}. (a) Results in GW and (b) on percent basis.
• Figure 5: System outcomes for the power and gas systems in 2050 where the load projection is based on an exemplary weather year. (a) Power capacity, (b) power generation. 'OCGT' and 'CCGT' are open and combined cycle generators, respectively. 'CCGT-CCS' is a 'CCGT' plant with carbon capture and storage technology. (c) gas consumption (top) and supply (bottom), and (d) annual power-gas system costs for exemplar weather year under different electrification scenarios, as described in Fig. \ref{['fig:demandresults']}, and decarbonization targets. The investment costs are annualized. In the upper legend, "Li-ion" is short-duration battery storage. In the lower-right legend, "VOM" and "FOM" are variable and fixed operating and maintenance costs, respectively. "CCS" is the cost of establishing carbon capture and storage infrastructure. 'Power gen + storage fix' is the investment and FOM costs for storage and power plants. 'Network Expansion' is the cost of establishing new transmission lines.