Short answer: no, unless for houses or flats that only need a tiny amount of heating energy.

Motivation: find out how solar (and wind) energy production match with typical household energy consumption in Belgium.

Solar energy production is often compared with household electricity consumption. At higher latitude this comparison is not accurate. Most solar energy is generated in summer and most household energy consumption is in winter. Also electricity is only a part of household energy. We illustrate this for Belgium - at a latitude of 50 degrees - in following parts:

• The first part shows the average household energy consumers in Belgium. Using data from Eurostat; we show that space heating consumes most energy.
• Part two shows the production/demand mismatch of solar energy for space heating. We also show that seasonal fluctuations of wind are a better match for heating.
• Part three models seasonal energy production and heating consumption for different mixes of solar and wind. Even when overdimensioning solar by a factor of two there is insufficient energy for heating in the winter season. Overdimensioning wind with a factor of four still results in a quarter of days where insufficient energy is available for heating. Taking January 2020 as an example we see that solar energy is low for the entire month - though solar has the highest monitored peak power - and in this period also wind energy production can be low for several days.
• Finally part four shows the share of fossil fuels for heating, as well as their yearly totals. Even though Belgium is a small country, the energy quantities are significant when considering storage.

The data shows that heating is a challenge both due to the magnitude of required energy and the production/demand mismatch with solar energy.

1. Household energy consumption in Belgium

Table 1 shows the end-user energy for Belgium from 2017 to 2020, as provided by Eurostat[1]. This includes home applications and does not include transportation. Clearly space heating is the biggest part, using almost three quarters of the yearly household energy. This is slightly more than the EU average of 63.6% for space heating due to the northern location and buildings with a low insulation level. In Belgium, household energy uses approximately a quarter of final energy over a year[2]. Household electricity use is over five times less, at 4.81 percent of total final energy in 2018 in Belgium [2]. Hence, comparing renewable energy potential savings relative to household electricity consumption is misleading; since household electricity consumption only corresponds to a minor part of total energy demand. In this text we use data from 2018 from table 1, since there is not much difference with more recent values for Belgium.

Table 1. Energy end-use for households in Belgium, 2017..2020. Data source - Eurostat[1]

Household final energy share	Space heating	Space cooling	Water heating	Cooking	Lighting and appliances	Others
Belgium 2017	73.8%	0.1%	11.4%	1.7%	12.7%	0.4%
Belgium 2018	73.5%	0.1%	11.9%	1.7%	12.5%	0.9%
Belgium 2019	73.2%	0.1%	11.9%	1.7%	12.8%	0.4%
Belgium 2020	72.7%	0.1%	11.7%	1.7%	13.2%	0.6%

2. Solar energy production mismatch with household heating

Due to the high latitude of Belgium, more solar energy is generated in summer than winter, and more space heating is needed in winter than in summer. This is shown in figure 1. We model solar production and space heating demand as follows:

• For solar data we use bias corrected satellite data from Pfenninger and Staffel[3][5] and computed and normalized hourly data for years 2005 up to 2015 [8].
• As an indicator for heating we compute degree-days with a base temperature of 15.5 degrees Celsius, using NASA MERRA-2 population weighted hourly temperatures [5][6] for years 2005 to 2015 [8].

Figure 1. Belgium - monthly percentage of solar, wind, heating and cooling degree days (year is 100%).

For the months January, February, November and December the average heating degree days are about 61 percent, while solar generation is 11.7 percent of the yearly total. For wind power about 43.2 percent of energy is produced in these months, which is better matched to winter heating energy demand.

Other models also point to a production/demand mismatch between solar and heating:

• Using the PVGIS 5.1 simulator [7] for Brussels [9], we obtain comparable results: 61 percent of heating degree days and 15.2 percent of solar generations for January, February, November and December. This assuming fixed solar panels facing south, with an optimal inclination and no shadowing effects in winter time.
• Using solar and wind energy forecast data of Elia for the years 2013 to 2019 [10], summarized in table 2. For the months January, February, November and December we obtain following average percentages of yearly production: 11.3 percent solar, 45.8 percent onshore wind and 42.3 percent offshore wind. These values are similar to those obtained for the Renewables ninja models, even though the year span is different. To estimate heating degree days for these years, and we obtain comparable results, specifically 60.5 percent of yearly degree days in the selected cold months.

Item	Average 2013-2019	2019	2018	2017	2016	2015	2014	2013
Solar	11.3%	12.4%	12.2%	10.1%	12.1%	11.3%	10.9%	9.88%
Onshore wind	45.8%	40.6%	46.8%	38.6%	45.9%	47.8%	54.0%	44.8%
Offshore wind	42.3%	39.4%	45.6%	40.5%	43.4%	45.8%	49.1%	39.9%
HDD1	60.5%	61.2%	63.8%	65.7%	59.3%	53.8%	64.0%	%57.3

_{(1) HDD = heating degree days with a balance point of 15.5 degrees Celsius, as obtained from Wolfram alpha for the EBBR weather station}

Based on these models we find that space heating demand is about 4 to 5 times larger in winter than what is produced by solar. And taking into account the space heating is the dominant energy consumer it is clear that there is a significant mismatch between solar energy production and household energy consumption.

3. Using renewables for a small heating load in Belgium

To illustrate the impact of the solar heating demand mismatch we consider the average natural gas consumption of a typical household ; 23.260 kWh when gas heating is used [11]. This load is significant, corresponding to 10 kWh of energy per heating degree day (Figure 1). In Belgium the maximum residential rooftop solar capacity is restricted to 10 kWpeak. This capacity is not sufficient to heat a home directly with electric space heaters. We can argue that additional insulation or a heat-pump can reduce this. To keep our analysis simple we assume a heat pump with a constant COP (coefficient of performance) of four, resulting in a heating load of 2.5 kWh per degree day.

This assumption is a simplification which is more benign for solar than in reality, as shown in the work of G. Emmi et al. [12]:

• The COP decreases in colder winter months, where it is more challenging to obtain a COP of four.
• Since Belgium has a cold climate thermal drift is an issue, since heating degree days are high while cooling degree days are small. A possibility is to inject solar heat from summer when a ground source is used for thermal heat[12]. Still electricity is required to use the heat pump in winter and we assume that it is proportional to the amount of heating degree days.
• Our model assumes that daily energy is available when needed during the day. Typically homes are heated in the morning, before the sun is available in winter. We assume that daily solar energy is available already in the morning when needed, assuming that a storage capacity of one day of energy is sufficient to shift demand and effects of a time shift are averaged out over a long period.

Using these assumptions we obtain a yearly heating energy demand of 5,840 kWh, one fourth of an average home in Belgium due to the assumption of a COP of four. Next we opt to overdimension the available solar and wind energy to 10 kWpeak. Combining this with the Renewables ninja data from Figure 1 we obtain yearly energy values as shown in Table 3. Ranging from about 1.75 times to 3.92 times the required energy for heating.

Energy mix	kWh/year for 10 kWpeak	yearly kWh surplus
100% solar	10,200 kWh/y	4360 kWh
75% solar, 25% wind	13,400 kWh/y	7560 kWh
50% solar, 50% wind	16,600 kWh/y	10760 kWh
25% solar, 75% wind	19,700 kWh/y	13860 kWh
100% wind	22,900 kWh/y	17060 kWh

Figure 2 shows a box plot of the energy surplus of produced daily energy for different energy mixes, with daily heating load subtracted. For solar we see that the average and medium surplus are negative for solar in the months November, December, January and February, even with a yearly surplus of 10,200 - 5,840 = 4360 kWh. Hence we can conclude that solar energy is not suitable for heating in these four cold months. Opting for 10 kWpeak into a mix of 75% solar and 25% wind improves the averages significantly, and this continues to improve until 100% wind. Still, even as 100% wind has a large surplus, insufficient energy is available about 25% of the days in the period from November till February.

Figure 2. Energy surplus in Belgium for different solar and wind energy mixes under a heating load and assuming ideal daily energy storage.

Another representation of wind and solar fluctuations is shown in Figure 3, without considering heating load. This shows daily data from Elia for the monitored capacity by in Belgium for January 2020[10]. The lack of solar energy is easy to understand, since it has the highest peak capacity (3887 MegaWatt-peak) and lowest amount of produced energy (81.7 GWh). Offshore and onshore wind produce much more energy even though they have a lower peak capacity. Still significant time-spans with low energy production can exist, as for example the period from 20 to 25 January. Heating demand cannot be time shifted for such long periods, so either energy storage or backup power is needed.

Figure 3. Wind and solar energy fluctuations in January 2020 for Belgium.

We observe that solar power is insufficient for the considered small heating load. Reducing the heating load or increasing installed solar power can help, still the production/demand mismatch will continue to exist in these cases. Specifically a large energy surplus of solar in summer and a tight solar surplus in winter. Another possibility is seasonal storage for solar. Still at present there do not seem to be viable solutions, as detailed in[13]. Even a 1,250 kWh battery at 100 Euro/kWh would result in a cost of 125,000 Euro storage cost, which is not viable. Using power to gas can result in a lower storage cost, still for heating application it will be favorable to use power to gas first for wind which is better matched with the heating load.

4. Household heating energy mix in Belgium

In previous parts we considered relative comparisons to illustrate production/demand mismatch between solar and heating, as well as seasonal fluctuations for a single home use case. In this section we show absolute energy numbers for household and heating energy. The next tables summarizes the energy mix for space heating. Table 4 shows that fossil fuels account for over 85 percent of total heating energy. In the category renewables and bio-fuels, using wood is the largest energy source for heating. So over 95 percent of heating results in direct CO₂ emissions.

Table 4. Energy mix for space heating of households in Belgium, 2017/2018. Data source - Eurostat [1]

Space heating fuel source	Electricity	Oil and petroleum	Natural gas	Solid fossil fuels	Renewables and bio-fuels	Derived heat
Belgium 2017	3.2%	38.0%	47.4%	1.4%	10.1%	0.0%
Belgium 2018	3.2%	37.9%	47.1%	1.2%	10.4%	0.3%
Belgium 2020	3.0%	41.5%	44.1%	0.6%	10.5%	0.2%

Table 5 shows the percentage used for heating per household energy source, and also quantifies household and heating energy in TeraWatt-hour (TWh). We see that about 12 percent of electricity is used for space heating, whereas about 89 percent of all other household energy sources are used for heating. Total heat energy for households in Belgium corresponds to 69.2 TerraWatt-hour for the year 2018, over four times as much as the household electricity used for other purposes than space heating. About 60% of this energy, 41.5 TWh, is consumed in the months January, February, November and December. Assuming that this can be reduced with a factor four still results in about 10 TWh energy consumption in these months. Suppose that we want to generate 10 TWh with solar during summer and to store it for winter at a cost of 100 Euro per kWh then we need a storage investment of 1000 billion Euro.

Techniques that reduce heating footprint, such as better isolation for example, will be more cost effective than storing solar energy for winter. Cheaper storage solutions exist [13] but are less suited for seasonal storage of solar power:

• Thermal solutions exist that store energy at a lower cost, but these result in significant losses for longer periods.
• Power to gas offers lower storage costs at the cost of more energy losses. To keep energy prices in winter affordable it would imply that solar power in warmer periods is cheap to compensate the losses in the power to gas conversion, and this for sufficient quantities for heating. Since little solar power is available in winter it makes more sense to first use power to gas for wind energy for heating.

Table 5. Energy mix summary of household energy in Belgium, 2018. Data source - Eurostat [1][2]

Belgium 2018	Household percentage	Household energy	Heating energy	Percentage used for heating
Electricity	19.5%	18.4 TWh	2.21 TWh	12.0%
Rest	80.5%	75.8 TWh	67.1 TWh	88.5%
+Oil and petroleum	30.3%	28.5 TWh	26.2 TWh	92.0%
+Natural gas	41.0%	38.6 TWh	32.6 TWh	84.4%
+Solid fossil fuels	0.8%	0.8 TWh	0.8 TWh	100%
+Renewables and biofuels	8.2%	7.68 TWh	7.2 TWh	93.7%
+Derived heat	0.19%	0.18 TWh	0.18 TWh	100%
Total	100%	94.2 TWh	69.2 TWh	73.5%

5. Conclusion

For Belgium solar is not suitable for meeting typical household energy demand. Space heating consumes almost 75% of household energy and over 60% of heating energy demand is in the period of November till February. In this period solar produces only about 10% of its yearly energy and seasonal storage is not yet economical. Wind energy is better matched with heating demand in Belgium and heat-pumps can lower the energy demand. Still the amount of energy needed for household space heating is significant and space heating is also needed for companies, schools and hospitals for example. Hence efficient heating of buildings using other solutions than solar energy will be required to lower the carbon footprint in a meaningful way.

And what about other countries? Countries closer to the equator will in general require less space heating, and the seasonal fluctuations of solar power will be lower. For hotter countries space cooling may be needed instead. The inverse is true for countries further from the equator. There even less solar energy will be available in the heating season.

Acknowledgement: many thanks to:

• Eurostat, part of the European Commission, for allowing free re-use of energy balances and related data. Website: ec.europa.eu/eurostat/web/energy.
• the Renewables.ninja team for open access to the v1.1 data-set of Europe, covering 1980-2016 data for weather, solar and wind on a country basis. Website: www.renewables.ninja. Version 1.1 of the original data-set is licensed under CC BY-NC 4.0.
• the PVGIS team for open access to PVGIS 5.1 data from 2005 to 2015. Source - "Photovoltaic Geographic Information System (PVGIS)", European Commission, Joint Research Centre (JRC). Online tool available at re.jrc.ec.europa.eu/pvgis.html - PVGIS Ⓒ European Communities, 2000-2020.
• Elia in Belgium for open access to solar and wind grid data at www.elia.be/en/grid-data. This data is licensed under the Elia Open Data License, which uses the CC BY-4.0 public license, and is governed by Belgian law.

Disclaimer: This work provides a summary of the original Eurostat energy data, converting ktoe (kilo ton oil equivalent) values to other energy units such as relative percentages and energy expressed as TeraWatt-hours (TWh). Values are rounded to 3 significant digits. Eurostat is not responsible for data conversion or rounding errors in doing so.

References:

History:

March 2023 August 2022 February 2022 November 2020