Elsevier

Renewable Energy

Volume 108, August 2017, Pages 169-178
Renewable Energy

An economic assessment of distributed solar PV generation in Sweden from a consumer perspective – The impact of demand response

https://doi.org/10.1016/j.renene.2017.02.050Get rights and content

Highlights

  • A household PV and demand response model were developed.

  • Electricity pricing schemes impact on PV size and demand response were investigated.

  • Net metering results in largest PV installation sizes.

  • Tax reduction scheme retains an incentive for demand response.

  • Hydronic demand response show significant impact on PV installation size.

Abstract

We present an economic assessment of the impacts of Demand Response (DR) and pricing schemes on the conditions for distributed solar photovoltaics, with the focus on individual households. An optimization model has been developed that minimizes the electricity cost for individual households with the option of dispatching DR loads. DR of appliances and hydronic heating (electrical water heating for both space heating and hot water) are investigated, as well as the effects of applying a monthly, hourly or net metering pricing scheme for selling excess generated electricity and a tax reduction scheme for electricity sold to the grid.

We show that for Swedish conditions a monthly net metering pricing scheme would result in the largest PV installations per household (median rated capacity of 4.2 kWp/household). The use of the tax reduction scheme reduces the installation per household (2.1 kWp/household), but with an installation being profitable for a larger fraction of the households. Furthermore, the tax reduction scheme retains an incentive for engaging in DR. The use of hydronic DR can support the same installations sizes as the tax reduction scheme, whilst Appliance DR is shown to have only a low impact on the profitability of a PV installation.

Introduction

The installation of distributed solar photovoltaic (PV) generation in Sweden doubled for the fourth year in a row during 2014 [21]. The future expansion of distributed PV will be dependent on the pricing scheme applied to electricity sold to the grid as well as the ability to use generated electricity locally in such installations. This work investigates the potential role of PV installations from the perspective of households, and investigates how electricity pricing schemes and demand response (DR) can either facilitate or hinder the implementation of PV for the private consumer. The analysis is made for the Swedish residential sector, with the aims of arriving at general conclusions that are applicable also to other markets.

The economic performance of a small-scale Distributed Solar Photovoltaics (DSP) installation, such as that operating at the household level, is dependent upon the consumers' payment scheme. In general, there is a substantial difference between the wholesale price of electricity that is sold to the grid and the price of electricity that is bought from the grid, as the buying price usually includes various transfer fees and taxes. Consequently, the household load profile and the size and generation profile of any DSP installation as well as the payment scheme for buying electricity are decisive factors in determining the economics of exchange with the surrounding electricity grid. While the current most common payment scheme for buying electricity for Swedish households is a fixed flat rate, which can be based on anything from monthly averages of the spot market prices to fixed rates over longer time periods of up to years [37], any excess generated electricity is fed to the grid and traded on the spot market where payment is cleared on an hourly basis. This restricts DSP installations to rather low capacities to avoid selling electricity to the grid, as the price received for sold electricity is too low to justify a larger investment. Starting in 2015, Sweden has implemented a tax reduction, in the form of a tax credit, of €0.069/kWh for excess electricity production that is fed into the grid from DSP [27]. Furthermore, a renewable energy certificate scheme is currently in place in Sweden, jointly with Norway, with an annually increasing quota of renewable electricity out of the total electricity generation [25]. Certificates are traded on a market in which excess certificates can be sold to non-renewable electricity producers, thereby providing economic incentives for renewable electricity generation. However, those households that have participated have only received certificates for the amount of electricity they have sold and not for their total generated electricity [43]. Investment support has also been adopted in Sweden and has a budget of €78 million for the period 2013–2018.

An alternative strategy to increase the value of DSP-produced electricity is to increase self-consumption of generated electricity, using Demand-Side Management (DSM) measures in the form of Demand Response (DR), i.e., shifting loads to hours of excess generation based on electricity price signals.

Previous related studies have investigated different financial support or electricity pricing schemes designed to facilitate the expansion of distributed power [13]. reviewed 16 studies concerning the costs and benefits of distributed solar PV in the USA. They conclude that there are many factors to include (e.g. environmental, grid support value and financial value) as well as different stakeholders to which the different benefits and costs may accrue. They identify several gaps in research, one being the need to investigate PV in the context of DSM [46]. calculated the social welfare generated (here defined as the sum of the consumer surplus, profit for the electric utility, and environmental benefits) by a net purchase and sell, a net metering system, and a feed-in tariff scheme. Yamamoto concluded that a net purchase and sell, as well as net metering would result in the highest level of social welfare provided there would be a sufficiently strong reduction in electricity consumption as a result of implementing the schemes. If the reduction in electricity consumption due to the measures turned out to be low, feed-in tariffs, which carry no incentive to reduce electricity consumption, would generate the highest level of social welfare [2]. compared the impact of time-of-use electricity rates and flat electricity rates on solar PV investments in California. Time-of-use rates charge higher prices for electricity during peak demand hours and lower prices during off-peak hours. Borenstein concludes that time-of-use pricing is beneficial for a vast majority of households with a PV investment [3]. compared renewable energy certificates with other policies, e.g. net metering and tax credits, with respect to their impact on PV investments in different US states. They show that RECs can be a more powerful driver in the implementation of solar PV than the other policies investigated due to their disconnection from the electricity price. However, the uncertainty of future prices for RECs is a major drawback for this scheme. In order to lessen the uncertainty they propose a guaranteed minimum price for the RECs. Several studies have investigated the potential of net metering compared to that of other financial support schemes [4], [5], [7], [10], [24], [32]. For instance [7], examined 200 households in California and concluded that net metering was preferable for the households over a feed-in tariff or hourly/monthly payment for excess electricity, with the caveat that this was based on the Californian tiered electricity pricing structure. For a Cypriot household [32], concluded that at the existing price levels, net metering was preferable to feed-in tariffs for a household with installations that exceeded 2.5kWp. However, all these previous studies applied a top-down methodology, with the exception of [7] and [32]; and investigated financial measures, while none of the studies included DR measures as alternatives to support the employment of DSP.

The combination of DR and distributed PV in a Swedish context has previously been investigated by Ref. [45]. Their aim was to investigate storage and the ability of DR to maximize self-consumption of generated PV electricity. They concluded that for low-excess generation, load shifting is preferable, whereas at high levels of PV electricity generation storage is better. This is because the drawback of losses from storage becomes small at large storage sizes and because for DSM, not enough energy is available to shift to hours of excess generation. However, no economic considerations were taken into account in that study [44]. also investigated the value of DR of appliances together with PV in Swedish households and concluded that savings of up to €20 per year are possible. However, the savings are dependent on the electricity price during the year investigated [31]. investigated different net metering and feed-in-tariff schemes together with DR, and they concluded that the economic savings for households were modest, whereas revenues for retailers and distribution system operators were reduced. However, if constraints were introduced into the electricity network, the distribution system operator could benefit. Castillo-Cagigal et al. [6] studied DSM and storage coupled with PV in an actual energy/electricity self-sufficient house. They showed that DSM could reduce the required storage size and increase self-consumption. However, they did not provide any economic evaluation of the profitability of DSM.

Although the above-mentioned studies provide valuable insights into the benefits of different pricing schemes, retail price structures, and DSM measures, there is a general lack of studies that compare DSM measures (in our case, in the form of DR measures) to pricing schemes with respect to the ability of DR to promote the implementation of DSP. Studies that investigate these aspects for Swedish conditions, in which there are different generation and usage patterns compared to continental Europe, are also lacking.

This study aims to investigate the potentials of DR and electricity pricing schemes as means to support the implementation of distributed PV in a Swedish context. The analysis is based on the electricity demand for single-family dwellings in Sweden.

Section snippets

Methodology

This study applies modeling of the hourly energy balances of individual households with the objective of minimizing annual costs of electricity. This is done by assessing several different cases in which factors that influence the economic performance of solar PV installations at the household level are varied, e.g., market setup, and whether or not loads can be shifted in time to exploit the generated PV electricity. Thus, this requires an analysis that includes hourly electricity generation

Cost of electricity and break-even size of PV installations

Fig. 3a shows the annual electricity cost during Year 2007 for all households when there is no investment in PV panels. Fig. 3b shows the median and range for the break-even ALR for the Swedish year 2007, i.e., the ALR at which the annual electricity cost combined with the annualized PV cost is equal to the annual electricity cost when not investing in PV panels, for five of the investigated pricing schemes. The yearly electricity cost is in the range of €620–€4050/year, with the upper end of

Discussion

The extent to which each investigated DR measure supports the installation of distributed PV depends on both the ability of the measure to lower costs and the capacity of the building stock to apply the measure. There are presently about two million single-family houses in Sweden, all of which have the theoretical possibility of implementing any pricing scheme or Appliance DR [41]. Around 300,000 Swedish households have hydronic systems that could make use of hydronic DR measures for space

Conclusions

The impacts of different DR measures, pricing schemes, and electricity price on investments in distributed PV are investigated for Swedish market conditions using a techno-economic optimization model on 21 Swedish buildings. The model includes the DRs of appliances and hydronic loads.

We show that under the given Swedish conditions and with an annual average electricity price of €34/MWh, implementation of a monthly net metering scheme gives the highest optimal ALR at a value of 3.0, resulting in

Acknowledgment

This work is financed by the Chalmers-E.ON initiative and the research program Pathways to Sustainable European Energy Systems.

References (46)

  • B. Shen et al.

    The role of regulatory reforms, market changes, and technology development to make demand response a viable resource in meeting energy challenges

    Appl. Energy

    (2014)
  • J. Widén

    Improved photovoltaic self-consumption with appliance scheduling in 200 single-family buildings

    Appl. Energy

    (2014)
  • J. Widén et al.

    Options for improving the load matching capability of distributed photovoltaics: methodology and application to high-latitude data

    Sol. Energy

    (2009)
  • Y. Yamamoto

    Pricing electricity from residential photovoltaic systems: a comparison of feed-in tariffs, net metering, and net purchase and sale

    Sol. Energy

    (2012)
  • R. Bohlin

    Beskattning av mikroproducerad El m.m

    (2013)
  • S. Borenstein

    Electricity Rate Structures and the Economics of Solar PV: Could Mandatory Time-of-use Rates Undermine California's Solar Photovoltaic Subsidies?

    (2007)
  • J.A. Duffie et al.

    Solar Engineering of Thermal Processes

    (2013)
  • E.ON. Våra elnätsabonnemang för dig som mikroproducent, Accessed: 2017 20 February...
  • A. Goodrich et al.

    Residential, commercial, and utility-scale photovoltaic (PV) system prices in the United States: current drivers and cost-reduction opportunities

    Contract

    (2012)
  • L. Hansen et al.

    A Review of Solar PV Benefit & Cost Studies

    (2013)
  • M. Iqbal

    An Introduction to Solar Radiation

    (1983)
  • D.L. King et al.

    Photovoltaic array Performance model

    (2004)
  • M. Kolhe

    Techno-economic optimum sizing of a stand-alone solar photovoltaic system

    Energy Convers. IEEE Trans.

    (2009)
  • Cited by (28)

    • Estimating national and local low-voltage grid capacity for residential solar photovoltaic in Sweden, UK and Germany

      2021, Renewable Energy
      Citation Excerpt :

      Focusing on the longest feeder likely causes a bias towards voltage violations in our results. We are interested in the technical hosting capacity, and therefore exclude economic limitations [31]. In addition, we exclude the impact from harmonics and flicker since the impact on hosting capacity is generally small, unless the grid is weak [32].

    View all citing articles on Scopus
    View full text