Deficits do not seem to occur in households when expected

[mabijkerk]

When I open a blank nl2019 scenario, I set the condensing combi boiler (gas) to 100% for households.

I can get the demand curve of apartments before 1945: FEVER_DEMAND_CURVE(V(households_useful_demand_for_space_heating_apartments_before_1945)). The maximum value of that curve is 3006 MW demand in a single hour. With the number of units set to 540698, the maximum demand of single residence is 5.56 kW.

I would therefore expect deficits to occur when I set the heat_output_capacity of the households_space_heater_combined_network_gas to 5.5 kW or lower in the dimensioning section: https://beta.energytransitionmodel.com/scenario/demand/households_heating_order/capacities. However, if I set the relevant input households_heater_combined_network_gas_heat_output_capacity to 5.5 kW, no deficits appear in the deficit table. Only when I set the capacity to 5.0 kW, deficits appear:

This is relatively close to what is expected, but I'm still wondering why this difference occurs. During testing we also encountered an example where the difference was larger, but I can't reproduce it now. Maybe @redekok knows when this occured?

[redekok]

During testing we also encountered an example where the difference was larger, but I can't reproduce it now. Maybe @redekok knows when this occured?

I thought I did but I'm actually not able to fully reproduce this bug. When maximally installing air-sourced heat pumps and increasing the typical useful demand values to 500 kWh/m2 for each category, barely any deficits occur. Changing the installed capacity per heat pump unit doesn't change anything. I think it was a variation on a scenario like this.

[noracato]

The households_space_heater_combined_network_gas has its defer_for property set to 4. Roughly meaning it has 4 hours to catch up with any deficit that occurred. So it does not surprise that there is this small difference from 5.5 to 5.0 to when we really see the deficits.

Or is there more to investigate that I'm not seeing?

[redekok]

Ah interesting, I thought we were gonna drop that attribute. @mabijkerk what are your latest thoughts on the defer_for attribute?

[mabijkerk]

Thanks for pointing this out @noracato! I also thought that we would drop that attribute. We could instead set the defer_for to 0 for all heaters.

This still produces an unexpected result though. In a blank nl2019 with the combined gas combi-boiler set to 100% and defer_for to 0, the detached houses from before 1945 begin to get some deficits from a heat output capacity of 5.8 kW. The capacity below which deficits are expected to occur lies at 5.5 kW though:

DIVIDE(
  MAX(FEVER_DEMAND_CURVE(V(households_useful_demand_for_space_heating_detached_houses_before_1945))),
  V(households_useful_demand_for_space_heating_detached_houses_before_1945,number_of_units)
)*1000

5.466262439750358

This seems like a minor issue to follow up on at some point. For now I will make no changes. When we follow up on this issue we can decided whether or not to change the defer_for attribute to 0 or to remove it altogether.