Electricity demand is lower than production, but there is no import and there are no blackout hours

[DorinevanderVlies]

An external user drew our attending to some issues in this scenario.

The total demand for electricity is lower than the electricity production.

However, there is no import of electricity and there are 0 blackout hours. There is some export of electricity.

Adding a gas plant (for example Gas CCGT) does not result in extra electricity production even though according to the demand vs production there still is a deficit.

@jorisberkhout told this might be due to a small bug related to electricity demand input for Merit.

[antw]

I went through each week in the "Electricity production per hour" chart, and can't find any period where supply fails to meet (or exceed) demand, so as far as Merit is concerned there are indeed no blackout hours. The HV network node has the same amount of energy entering as leaving. Is there evidence of a mismatch between supply and demand in ETEngine, or is it just in the ETModel front-end (e.g. charts) where we see the error?

If the latter, is it possible we're simply missing a source of electricity from the electricity_production group? Or something amiss in one of the electricity_produced_from_(carrier) queries?

[DorinevanderVlies]

Is there evidence of a mismatch between supply and demand in ETEngine, or is it just in the ETModel front-end (e.g. charts) where we see the error?

I've not yet looked into ETEngine. So as far as I know it's just in the ETModel front-end.

If the latter, is it possible we're simply missing a source of electricity from the electricity_production group? Or something amiss in one of the electricity_producedfrom(carrier) queries?

I'll look into this

[antw]

@jorisberkhout told this might be due to a small bug related to electricity demand input for Merit.

Merit is told that total electricity demand is 30.729 PJ, and Merit meets or exceeds demand at all times (total production of all Merit order participants is 37.754 PJ). The chart shows output of under 30PJ.

That's not to say it's definitely a chart or query issue, but it would certainly be my first guess.

[jorisberkhout]

That's not to say it's definitely a chart or query issue, but it would certainly be my first guess.

You're probably right, @antw . I have attached an Excel-file to this issue that shows the production of all converters in this scenario (loads.308620.xlsx, see Sheet 2). Would you have time to analyse this, @DorinevanderVlies . Maybe you can figure out which production is missing form the chart.

[DorinevanderVlies]

I found that agriculture_chp_engine_network_gas and industry_chp_combined_cycle_gas_power_fuelmix seem to be missing from the chart.

There is also something with biomass being coal being 0.467 PJ higher in the excel file than in the chart. Solid biomass on the other hand is 0.467 lower. The weird thing is that I cannot find coal attributes that can make up 0.467.

Furthermore green gas is 0.0098 PJ in the production chart, but there is no green gas in Excel. The following shows my comparison of the ETM production chart and the Excel file from @jorisberkhout.

loads.308620 dv.xlsx

[jorisberkhout]

Good observations, @DorinevanderVlies ! I looked into this a bit more and found that not only are agriculture_chp_engine_network_gas and industry_chp_combined_cycle_gas_power_fuelmix missing from this chart, but if you query their electricity production in this specific scenario this returns 0 in both cases. My suspicion is that merit takes these converters into account, but afterwards their demands are set to 0 because the sectors they are in (agriculture and industry respectively) are excluded in this scaled scenario. What do you think about this suspicion, @antw ?

[jorisberkhout]

The other two observations are easier explained, @DorinevanderVlies .

In you Excel-analysis you labelled two cofiring converters as using just coal. These converters do, however, use 50% coal input and 50% solid biomass input. This explains why you see to little solid biomass consumption and to much coal consumption.

The green gas is a result of the fact that most gas-fired electricity producer use network_gas, which is a mixture of natural gas and green gas. The scenario has 0.4% green gas in the gas network which explains the 0.0098 PJ electricity production from green gas.

[antw]

My suspicion is that merit takes these converters into account, but afterwards their demands are set to 0 because the sectors they are in (agriculture and industry respectively) are excluded in this scaled scenario. What do you think about this suspicion, @antw ?

100% correct. Good catch.

For producers in the energy sector we'd set number of units to zero. This prevented their used in Merit. I've widened this to producers in all sectors:

The fix is now on beta; should we also apply it to production? I don't expect there to be any fallout.

[jorisberkhout]

Thanks for fixing this @antw !

The fix is now on beta; should we also apply it to production? I don't expect there to be any fallout.

I can imagine that there will be changes to the merit order results for some scaled scenario's now that electricity producers in the agriculture and industry sector are taken into account correctly. I do not know if this will cause any issues. I guess @DorinevanderVlies knows more about scaled scenario's. Could you decide whether this can be applied to production or needs some more testing, @DorinevanderVlies ?

[DorinevanderVlies]

Clever thinking @jorisberkhout. And good fix @antw.

Could you decide whether this can be applied to production or needs some more testing, @DorinevanderVlies?

Yes I will look into this. I'll try to do it today, but it will probably be after the weekend.

[ChaelKruip]

The fix is now on beta; should we also apply it to production? I don't expect there to be any fallout.

Let's do this first thing in January @antw! Our external user would like to use the current results on production for the coming week.

[DorinevanderVlies]

Could you decide whether this can be applied to production or needs some more testing, @DorinevanderVlies?

I looked into previously made local scenario's on the beta server and found no issues in them. So, I think this fix can be applied to production.

[ChaelKruip]

Closing this. Reopen if this is still relevant.