Variability in electricity consumption by category of consumer the impact on electricity load profiles Philipp Andreas Gunkel1 Henrik Klinge Jacobsen1 Claire -Marie Bergaentzlé1 Fabian Scheller12 and

2025-05-06 0 0 1.74MB 37 页 10玖币

侵权投诉

Variability in electricity consumption by category of

consumer: the impact on electricity load profiles

Philipp Andreas Gunkel1*, Henrik Klinge Jacobsen1, Claire-Marie Bergaentzlé1, Fabian Scheller1,2 and

Frits Møller Andersen1

1 Sektion for Energy Economics and Modelling, DTU Management, Technical University of Denmark, 2800 Kongens

Lyngby, Denmark

2 Faculty of Business and Engineering, University of Applied Sciences Wurzburg-Schweinfurt, Ignaz-Schon-

Street 11, 97421 Schweinfurt, Germany

* Correspondence: phgu@dtu.dk; Produktionstorvet 424, 2800 Kongens Lyngby, Denmark

Abstract. Residential electrification of transport and heat is changing consumption and its characteristics

significantly. Previous studies have demonstrated the impact of socio-techno-economic determinants on

residential consumption. However, they fail to capture the distributional characteristics of such consumer

groups, which impact network planning and flexibility assessment. Using actual residential electricity

consumption profile data for 720,000 households in Denmark, we demonstrate that heat pumps are more likely

to influence aggregated peak consumption than electric vehicles. At the same time, other socio-economic

factors, such as occupancy, dwelling area and income, show little impact. Comparing the extrapolation of a

comprehensive rollout of heat pumps or electric vehicles indicates that the most common consumer category

deploying heat pumps has 14% more maximum consumption during peak load hours, 46% more average

consumption and twice the higher median compared to households owning an electric vehicle. Electric vehicle

show already flexibility with coincidence factors that ranges between 5-15% with a maximum of 17% whereas

heat pumps are mostly baseload. The detailed and holistic outcomes of this study support flexibility assessment

and grid planning in future studies but also the operation of flexible technologies.

Keywords: residential electricity consumption; household characteristics; consumption distribution; peak;

electrification

1. Introduction

This study provides a comprehensive summary of residential data on electricity consumption at a level of detail

that makes it suitable for further studies and applications in research, public services and industry. Electricity

consumption is subject to change due to the electrification of heat and transport in the context of the green

transition [1]. In response to this development, several key areas in the energy sector, such as the generation of

electricity, network planning, grid tariffs and tax design, are being reconsidered [2–4]. Consumer groups with

2 of 37

different socio-techno-economic characteristics will therefore face changes in their electricity bills that are more

dependent on the timing, location, peak and distribution of their electricity consumption [4]. Distributional cost

effects across all socio-techno-economic groups are expected. Consequently, more detailed research on

residential electricity consumption, its distributional characteristics and developments is necessary to anticipate

future challenges such as network development and policy design.

Residential electricity consumption varies across socio-economic parameters and technical equipment.

Previous studies have demonstrated that it is mainly determinants such as the type of dwelling, the heating

system used and the charging of electric vehicles that significantly affect consumption levels and daily peaks

[5]. Further influences on consumption levels range from the number of occupants [6] to the number of

bedrooms [7], the dwelling area [8], the floor area [9], incomes [10] and the household's ownership of physical

appliances [11] and electric vehicle [12]. Occupant characteristics and living conditions, which are often

correlated with income, also play a significant role [13]. Individual profiles are analyzed by [14] to identify

different clusters of consumers based on socio-economic factors determined by survey data. A 3,326 smart

meter records dataset is divided into 6 clusters with different peak consumption. Socio-economic factors have

large effects on the association with clusters. However, households also showed the characteristic of moving

from one cluster to another depending on the season. Therefore, residential patterns regarding similar groups

are subject to changing patterns that confirms the need for a detailed and fragmented investigation of residential

consumption profiles, both between and within chosen socio-techno-economic groups, since electricity profiles

are heterogeneous [15]. Similarly, [16] clusters residential consumption of 5566 households in London to

investigate consumption behavior. In the end, the study results in consumer clusters with daily profiles that can

help retailers to optimize their market participation based on customer segmentation. [17] improved clustering

methods by focusing on behavioral characteristics of consumers changing their consumption pattern over time.

[18] further analyzes the change of residential electricity consumption by consumers adopting EV and PV. The

authors observed using a difference-in-difference method that demand changes due to new technologies and

behavioral changes. Most studies using clustering have a consumption-first focus. However, grid operators, grid

and city planners, and policymakers focus on a socio-economic-centric view. Consumer change clusters over

time while socio-economic categories are more stable and are relevant to calculate the impact of policy

initiatives and infrastructure planning [19,20]. Most research links individual electricity profiles and household

characteristics, ranging from analyses of large samples [5] via representative smart-metering surveys [21],

longitudinal cross-sectional data [22,23] and specifically collected information on residential groups [6,8–

10,13] to analysis of different temporal resolutions [24–26]. The studies cited utilizing common average- and

3 of 37

regression-based methods are useful in drawing unified conclusions and average relationships [22,27,28].

However, they deal with various parts of the conditional distribution identically and neglect its heterogeneity

[22]. Also, a comparable holistic analysis is missing as studies focus on subsets of categories or specific

technologies, which makes comparisons harder due to differences in time, location, temperature or behavior.

For specific hours or timeframes, quantile regression is a further option, as applied among others by [15,22,25].

To a certain econometrical degree, those values are useful in understanding consumption. Nevertheless, the

unavoidably high level of aggregation makes the data less suitable for estimating the named areas where timing,

location, distribution, outliers and noise are expected to be decisive. The investigation of individual profiles is

thus necessary due to their heterogeneity [6]. [29] presents a stochastic analysis of plugin and availability pattern

of 10 electric vehicle used in the service sector, while [30] analysis the charging behavior of 221 electric vehicle

for 78 days in the UK. Plugin pattern of electric vehicle have mostly been analyzed through theoretical and

survey approaches such as in [31,32]. Probabilistic heat pump pattern are statistically analyzed by [33] for 19

households for the month of January in Ireland. A main outcome of the study is a coincidence factor for the

usage of heat pumps follows a gamma distribution with a strong baseload with long and flat tails relevant for

network planning problems. Similarly, [34] shows considerable flexibility potentials provided by energy

communities using electric vehicles and heat pumps under uncertainty. However, a clearer view on realistic raw

data is required to allow for optimal scheduling of smart home systems that have to deal with uncertainties

related customer behavior in relation to charging and heating, but also production from renewables [35–38].

This will help to mitigate and avoid costly distribution grid reinforcement as shown by [39]. Future studies

therefore are in need of detailed insights into electricity consumption pattern across and within customer groups

including uncertainty to improve prediction and operational models.

To address many of the named shortcomings, this research aims to reveal fundamental differences in

residential electricity consumption between socio-techno-economic categories at the individual dwelling level

in Denmark. In contrast to previous studies, the analysis does not only focus on average and aggregated

consumption analyses. It applies median, variance, and probability approaches and thus the distribution of

electricity consumption, which is of great importance to get a sense of the marginal impacts across the individual

electricity consumption profiles. While existing articles provide insights about the average relationships of the

electricity demand and respective variables, we present distribution shapes of residential electricity

consumption profiles for socio-techno-economic consumer groups. This approach offers new insights into

significant uncertainties for flexibility measures and grid adaptations at the centers, tails and peaks of the

consumption distributions and their temporal appearances. Furthermore, we quantify electricity consumption

4 of 37

variations within existing household categories while providing the same general properties such as time,

location and climate conditions. In the end, this study provides a holistic and comparable analysis across socio-

economic groups and technologies. The analysis is thereby also highlighting the heterogeneity within each

group. Consequently, this study provides an analysis of residential electricity consumption that supports grid

planning and policy making further and provide a beneficial view for data science and scheduling applications

as well as offer direct comparisons between technologies such as heat pumps and electric vehicles that have so

far been studied only individually in literature.

Our approach shows that techno-economic determinants like heat pumps influence with higher probability

the aggregated peak consumption than electric vehicles while socio-economic determinants such as occupancy,

living area, and income show little impact. Although electric vehicles generally contribute with a lower

probability to residential consumption than heat pumps, in the event they significantly demonstrate a higher

magnitude for individual hours of the day. By demonstrating this, we rely, unlike other studies, not on smart

metering surveys and selective field trails from a limited number of households but on a unique and

comprehensive dataset from a large number of individual dwellings into account. The dataset covers

approximately 720,000 households with an hourly meter of 2017. The Danish Transmission System Operator

(TSO) Energinet collects data from all hourly meters and delivers it delivers to Statistics Denmark. The smart-

meter data is linked to administrative registers giving reliable information on household categories. Thereby,

the applied unique dataset from Statistics Denmark with the large amount of individual household data also

decreases the impact of the volunteer bias present in most studies in the literature.

The study is structured as follows. Section 2 introduces the Danish Energy system in general and then

presents a closer look on the residential sector as well as the data and material of this study. After that, Section

3 shows the applied methods and measures to compare residential electricity consumption across consumer

groups. Section 4 summarizes the main results and compares the influence of different socio-techno

characteristics on statistical measures and distributions of residential electricity consumption. The discussion is

performed in Section 5. Firstly the results are compared and validated with existing literature and generalized

to other geographical locations when possible. In the second part of the discussion the results are put into context

of network planning and load forecasting. The last subsection of the discussion outlines research potentials such

as grid tariff and tax designs. Section 6 summarizes and concludes the study.

2. Materials - Residential energy consumption in Denmark

5 of 37

This section gives at first a general context of the electricity sector in Denmark regarding its generation,

consumption and then focuses on the energy demand of Danish households. After that, the case study data is

introduced, covering the categorization of household archetypes by socio-techno economic factors. The last

subsection summarizes the methodology and presents the calculated indicators.

2.1. Danish electricity sector and residential consumption

The Danish electricity sector has undergone a change in production and consumption over the past

decades. Since the 1980s Danish energy production shifted towards the expansion of renewable energy

resources such as wind power due to the ban of nuclear power originating from strong public resistance [40].

The contribution of wind power to the total electricity production reached already 48% in 2017. Flexible

generators, mostly co-generation power plants supplying district heat, covered around 50%, whereas solar PV

only played a minor role [41]. Danish politics remain ambitious with their plans to meet the requirements of the

Paris agreement and are currently on track [42]. The Danish Energy Agencies foresees an expansion of solar

PV by 445% and 643% in 2025 and 2030, respectively, an onshore wind development from 4.4 GW in 2018 to

6.2 GW in 2030, and an offshore wind capacity of 5.6 GW, which represents an increase by 435% [43]. The

incoming variable resources cover the rising electricity needs from the electrification of several sectors,

particularly the residential, that is also supposed to serve the needed flexibility via demand response.

Unlike several other European countries, Danish electricity consumption is largely affected by residential

consumers. Figure 1 summarizes the division of electricity consumption by sector on a typical winter day. The

pattern of the industry sector mainly influences the total electricity peak in the morning, including some

contributions by the residential sector. Contrarily, the residential sector is clearly responsible for the late

afternoon peak.

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

Variabilityinelectricityconsumptionbycategoryofconsumer:theimpactonelectricityloadprofilesPhilippAndreasGunkel1*,HenrikKlingeJacobsen1,Claire-MarieBergaentzlé1,FabianScheller1,2andFritsMøllerAndersen11SektionforEnergyEconomicsandModelling,DTUManagement,TechnicalUniversityofDenmark,2800KongensLyngby,...

展开>> 收起<<

Variability in electricity consumption by category of consumer the impact on electricity load profiles Philipp Andreas Gunkel1 Henrik Klinge Jacobsen1 Claire -Marie Bergaentzlé1 Fabian Scheller12 and.pdf

共37页,预览5页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

Variability in electricity consumption by category of consumer the impact on electricity load profiles Philipp Andreas Gunkel1 Henrik Klinge Jacobsen1 Claire -Marie Bergaentzlé1 Fabian Scheller12 and

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: