In Belgium, residential heating needs are, on the average, of 15 to 18 MWh/y, while the electricity needs (excluding any heating) are rather of 2,5 to 4 MWh/y, on the average.

This means that the heating needs are roughly 5-6 times the electricity ones.  The heating needs are mainly driven by the thermal insulation of the dwelling and subsidiarily by the technic used to produce that heat.

In a previous article (https://edenergy.be/heat-pumps-paradigm-or-fallacy/?lang=en) we have seen that the effective actual COP of a (air/water) heatpump is rather around 1,5 (instead of the claimed theoretical value of 3 to 4, widely spread in the public).

This means that 15 MWh/y consumed by a high yield condensing gas boiler (the most current type sold for more than 10 years), cannot be reduced lower than 10 MWh/y, making use of an excellent air/water heatpump.

Nevertheless, in the practice, we observe that dwellings equipped with these heatpumps consume around 10 MWh/y of electricity (in total), thus including the 2,5 to 4 MWh/y to cover the other needs, so that the heating (the heatpump) consumption does not represent more than 6 to 7 MWh/y.  But these dwellings are much better insulated than the ones consuming 15MWh/y for their heating, obviously only 1/3 to 1/2 (achieving thus a reduction 50% to 65%) of the heating needs observed for the common dwellings.  This just to draw the full context.

So, even in the best cases, which are not representative of the average, the total electricity demand necessary to include the heating needs will be 3 to 4 times (300% to 400%) as high as the one of today, at household level, while the DGO’s already fear for the widespread of electric cars which, in turn, will only increase the electricity demand by 50% (0,5 time), thus at most 1/6 of the of the increase related to the use of heatpumps.

The 1^st reaction would be to ban (or forbid) the heatpumps, to preserve the electricity grid and avoid a huge increase in the grid cost (for its reinforcement/upgrade) that would be incurred by the generalisation of heatpumps, but lobbies in behind are probably too strong for that, I presume.

Now, if we can’t do otherwise than accept the heatpumps as a fact (or given), we can try to reduce its offtaken power.  The consumed energy (kWh/y) being determined by the dwelling insulation (for given outside and inside temperature profiles), it can’t be reduced, but well the offtaken power (kW, and particularly its yearly peak: max kW).

Nothing magic, the consumed energy over a year is equal to the sum over the year of the power (its 15’ average in Belgium), offtaken during each quarter of hour, multiplied by a quarter of hour:

The quarter of hour is used because in Belgium the electricity measurements are achieved on a quarter of hour basis (but is not so in all countries, e.g. in France the basis is 5’).

If we dispose of a heat storage, the heating power can be withdrawn from this storage, while the storage can be filled continuously at a defined (lower) power rate. At least, this is the hope we are investigating.

The heat storage contains an amount of energy, from which heating demand energy can be offtaken at a defined rate (dwelling heating power) and simultaneously filled at another power (heatpump heating power, thus with a COP of 1,5: 1,5 × heatpump [electrical] power).  The heating storage is nothing else than a thermal battery, in the practice, a very good insulated (buffer) vessel/tank.

The point is thus, how can we design the best couple heat storage (stored thermal energy) / heat filling rate (1,5 × heatpump power), to quench the heating needs of a dwelling.

The 1^st thing to do is to identify the highest heat demanding day, to get an estimation of the maximum daily heating needs.  If we then build a storage capable of storing this energy and simultaneously refill the store with a power equal to this energy divided by 24h, we will be able to 1) provide the heating necessary for any day (any day need ≤ highest demanding day need), and 2) refill it to its maximum during the same day (since the designed power = highest demanding day need / 24h).

But this design is probably not the optimal one.  If the storage capacity is increased, the refilling power can perhaps be reduced, and conversely if the storage capacity is decreased, the refilling power would increase.

Increasing the storage capacity will cope with dimensional constraints (no one can install a container in her/his dwelling), while reducing the storage will require higher (refilling) power, which will cope with grid (congestion and tariff) constraints.

Let us thus assess the actual value of both within the playing field for the couple storage (capacity)/(filling) power.

Of course, the size of the storage-power couple depends on the spread of the heating demand along the year (the demand curve).  Long series of days with high demand will have another impact than a more randomly distributed days with high demand. We have thus to begin with assessing the yearly daily-demand curve.  Fortunately, it is well known (at least among the people active in the residential heating business, thus around 0,5% of the population😊) that the heating demand is strongly correlated to the so called ‘degreeday’ (a concept taking into account the outside temperature, the heat exchanges and the thermal inertia, more details: https://www.synergrid.be/fr/centre-de-documentation/statistiques-et-donnees/degres-jours).

So that we can use the degreeday curve to profile the heating needs for any day:

The degreedays are (calculated and) published on a daily basis.

Now, each year is different. We have thus to select a one that was representative, thus not deviating too much from a normal (standard) year, like 2021.  Besides, the calculations were also made for 2022, which was globally a colder year, to get feeling of this colder aspect on the results.

With the heating (demand) curve (distributed along the year), we are able to start the calculations.  We will calculate the average over a period as well as the deviations around it, for each day along the period, and then so for each period.  The average will define the (refilling) power, and the (absolute value) of the maximum by day of the sum of the negative deviations will define the storage (capacity).  Really ? … well not exactly, both will be majorants (in absolute value) for the searched values, thus a kind of secure value guaranteeing the heating needs satisfaction on any day, but exceeding them, even on the highest demanding day.

The real values for any couple can only be obtained by iteration.   I haven’t found the mathematical demonstration of this statement, but as a man ever told me: the practice is not finding the explanation for something that works while the theory is having the explanation of something not working.  Let us keep on the practical side.

The highest demanding day for the 1day period gives the maximum power that could be necessary without any storage (or utmost the energy demand for this day).  Starting from values calculated 7 days periods, for both years 2021 (globally near a standard year over the 3 first and 3 last months) and 2022 (globally warm over the 3 first and 3 last months), iterations have been run on 4 to 5 storage sizes in order to determine the related necessary power.

It is noticeable that even 2022 has been globally warmer than 2021, the averages are some % above for this year, while the averages by day (of each period) of the sum of deviations is smaller by 30 to 40% for the warmer year, showing that the distribution of the heating needs over a year are more determinant in sizing the heating system than the global values.

A warm year has an influence on the total year heating needs as well as its distribution along the year.  Hence a standard year is characterized by 2.252 degreedays, we have to apply a rule of 3 on the yearly standard heating needs to get the applicable yearly one.  For 2021, the rule of 3 gives a factor of 1,01 (thus the heating needs are of 1,01 × 16 MWh/year, thus 16,2 MWh/year) while for 2022 this factor is of 0,85 (thus 13,6 MWh/year).  These values will be distributed along their respective year.

Now, we have to translate the obtained couple of values storage (capacity) / (refilling) power into understandable terms.

We can see that even to save only 7 to 8% on the heating power, the necessary storage capacity reaches directly more than 700 litres (respectively 677 for 2021 and 626 for 2022), showing also that a warmer year does not significantly impacts the necessary storage size.

Further 700 litres of water weights 700 kg, on a surface of roughly 0,5 m², thus an equivalent charge of more than 1.400 kg/m² while the dwelling floors are commonly calculated to support 300 kg/m².

The conclusion is that, using a storage based on sensible water heat, to reduce the maximum offtaken heating power is not possible, even for small reduction of this power (less than 8% as presented above).

Of course, they are other types of heating storage, that could perhaps lead to less volume and weight, but today they are not available on the market.

This means that the heatpumps will require much larger maximum grid offtake and thus require huge investments on the grid, if they would be generalized.  These investments will be translated into huge grid costs increase, charged to every grid user, while putting the reinforced grid at its limits at specific moments.