buildings, agricultural CHPs: Mismatch in Full Load Hours (ETM) and Full Load Hours (profile)

[Richard-Deuchler]

buildings_collective_chp_natural_gas have a Nominal electrical capacity of 0.466 MW. The profile that we use in the merit order has a Full Load Hour characteristic: 2,680 h. This is smaller than the Full Load Hours specified in the ETM backend: 3,942 h. As shown in profile specifications, this is not 'allowed". In consequence, the buildings CHPs will actually produce up to 0.685 MW within the hourly calculation in merit order, which is above their nameplate capacity!

Problems:

• my evaluation of excess electricity is not transparent. I actually produce more must-run power than what the installed capacity would allow for.
• the merit order works with a residual demand that is too low in winter (considering all 870 buildings_collective_chp_natural_gas units, the residual demand may be 'off' by about 200 MW for many hours of the year)

Possible solutions:

• ignore the issue
• adapt Full Load Hours in the ETM to the Full Load Hours of the buildings profile (I will do that for my research Mathematica)
• find a new profile for buildings_chp that has a different shape

Note:

• the agricultural CHPs have the same problem, but it leads to a smaller mismatch, since FLH of profile and ETM are very similar. (error of ~60MW for all agricultural CHPs)
• the industry CHPs do not cause any trouble, because the FLH (profile) = 8760,

[Richard-Deuchler]

I found more participants (not installed by default). However, some of them may cause severe trouble. Keep in mind: intrinsic Full Load Hours of the buildings_chp: 2,680 hours

All of these participants are triggered by the same buildings_chp profile: ("up to" refers to FLH that depend on the demand. The other FLH are fixed)

Should the user decide to vote for a lot of buildings_collective_chp_wood_pellets, the residual demand curve might get severely messed up.

[Richard-Deuchler]

including @ChaelKruip

[ChaelKruip]

Also interesting to see that the fuel_cell runs for more than five year per year!

[ChaelKruip]

Also see https://github.com/quintel/etengine/issues/525

[AlexanderWirtz]

I believe this is a modeling issue related to number of households. assigning to @ChaelKruip as I have an idea what these devices should do, but not why this happens

[Richard-Deuchler]

to come to a solution for this ticket: the current buildings_CHP profile has only 2679 full load hours. This makes sense for gas boilers, which have a very high capacity. Gas boilers are switched on/off frequently, which explains the low FLH. New technologies, like heat pumps and fuel cells but also central CHPs, have much higher FLH because they operate closer to their peak load most of the time.

would it be an idea to define new profiles for all heating technologies that

• have a low capacity and therefore run more or less constantly (e.g. micro-chp, fuel cell)
• central CHPs, which do not behave like distributed gas boilers

We could just use a flat profile like industry_CHP, or define new ones. I also understand that we have a new intern starting soon, who is going to work on the heating of buildings/households? Maybe he will have much better ideas.

[AlexanderWirtz]

@Richard-Deuchler yes. I finally understand this issue now. Sorry, bit slow. We should definitely do this. CHP profiles exist to group CHPs together by their type of use, not by their sector. base-load technologies need other profiles than technologies that are switched on and off all the time.