Page 1 of 1

Nuclear power causes 38 GW power draw on system

Posted: Mon Sep 06, 2021 12:44 am
by FizwizDeadguy
I built 2 nuclear power supply plants, each consisting of 6 power reactors and their required supporting structure. I did each one at a time. I had no issues. Sometime later I built a 3rd plant, using the same blue print I used for the first two. With the structure partially completed I noticed I had no power on my electrical grid. I showed 0 satisfaction and a demand of 38GW and all the turbines were not producing. My demand was 1.2 GW before building the third plant. Removing the plant brought my grid back up. Laying the blue print down crashed it again. My demand list did not show anything asking for gigawatts of energy. To get every thing working again, I had to disconnect all 3 plants from my grid, at which point the demand dropped back to 1.2 GW. I connected the 3 plants back in, and everything is still working. Any ideas what might cause this?

Re: Nuclear power causes 38 GW power draw on system

Posted: Mon Sep 06, 2021 7:11 am
by Laie
Every consumer has a small internal buffer for electricity. Not much, just enough to run for a fraction of a second at full power.

Your demand as displayed in the power tab is how much is missing from all consumer's buffers at the end of a tick. In ordinary times, that is the amount of power you have consumed during that tick. However, if you are in a brownout/blackout, these buffers begin to run empty, and that shows up as ludicrously high demand.

In first approximation, the "demand" you see in a blackout is the maximum your factory could conceivably consume, if every machine was running, every inserter swinging, every roboport charging 4 bots, and so on. That has nothing to do with your third nuclear plant, it is an artifact of being in a blackout. I guess you somehow cut your power when placing the third blueprint? Though I wonder how that could happen.

Re: Nuclear power causes 38 GW power draw on system

Posted: Mon Sep 06, 2021 6:46 pm
by mrvn
Note that if the demand from blackout is 38GW and you only produce 1.2GB then it takes over 30s for the internal buffers to fill assuming none of the entities do any work. If they consume 1GB and you produce 1.2GB after blackout then it takes over 3 minutes for the internal buffers to refill. The less surplus energy you can produce the longer it takes for the 38GW demand to sink to your normal level. But you should see it shrink over time, it shouldn't stay at 38GW and then suddenly drop to 1.2GW when the buffers are full.

I assume you build a 3rd plant because you run out of energy or nearly so. Was your 3rd plant heated up already when you connected it? If not then it probably contributed nothing to the supply yet and the first 2 plants where simply not up to filling the internal buffers after blackout in a quick fashion. But after you tried to fix that for some time the 3rd plant has heated up so now you have a lot more power available.

Re: Nuclear power causes 38 GW power draw on system

Posted: Mon Sep 06, 2021 8:18 pm
by FizwizDeadguy
In first approximation, the "demand" you see in a blackout is the maximum your factory could conceivably consume, if every machine was running, every inserter swinging, every roboport charging 4 bots, and so on. That has nothing to do with your third nuclear plant, it is an artifact of being in a blackout. I guess you somehow cut your power when placing the third blueprint? Though I wonder how that could happen.
I assumed I had cut the other two plants out of the network when I put down the 3rd, but I still have my old 'steam plant' connected. It and a small remnant of my solar grid. They did not provide any power to the system at the time either. And cutting the first two out of the system is what let it reset, and show it only needed 1.2 GW.
I assume you build a 3rd plant because you run out of energy or nearly so. Was your 3rd plant heated up already when you connected it? If not then it probably contributed nothing to the supply yet and the first 2 plants where simply not up to filling the internal buffers after blackout in a quick fashion. But after you tried to fix that for some time the 3rd plant has heated up so now you have a lot more power available.
I actually still had quite a bit of excess power (.4 GW) I built the new plant because I was getting ready to build a massive mining and manufacturing facility for blue circuits with full modules and beacons. Between that and a sudden influx of laser fire, I didn't want to take any chances. As for the 3rd plant heating up, I now wonder if the reactors', exchangers', or turbines buffers' (assuming they also have them) were somehow the issue, since the reactors were not heated (or even fueled yet) when the bots started to fill in the plant with pieces.

Re: Nuclear power causes 38 GW power draw on system

Posted: Tue Sep 07, 2021 1:39 am
by mrvn
Oh and don't forget that construction bots need to recharge using up power. And roboports are I think the worst entities after a blackout because they have HUGE energy buffers that need to be charged again. Your roboports combined might well eat up 30+GW if they all charge at the same time. After that they only need as much as your bots consume, which is usually way less.

Re: Nuclear power causes 38 GW power draw on system

Posted: Tue Sep 07, 2021 4:17 pm
by Laie
mrvn wrote: Tue Sep 07, 2021 1:39 am roboports are I think the worst entities after a blackout
Quick check, a lone roboport on an isolated grid reports a demand of 5 MW, a laser turret is asking for 9.6 MW.

Re: Nuclear power causes 38 GW power draw on system

Posted: Wed Sep 08, 2021 4:08 am
by Silari
Laie wrote: Tue Sep 07, 2021 4:17 pm
mrvn wrote: Tue Sep 07, 2021 1:39 am roboports are I think the worst entities after a blackout
Quick check, a lone roboport on an isolated grid reports a demand of 5 MW, a laser turret is asking for 9.6 MW.
Beacons are terrible, IIRC, and possibly the worst one. Especially since late game bases tend to have so many.

Re: Nuclear power causes 38 GW power draw on system

Posted: Thu Sep 09, 2021 1:55 pm
by Tertius
Silari wrote: Wed Sep 08, 2021 4:08 am Beacons are terrible, IIRC, and possibly the worst one. Especially since late game bases tend to have so many.
Usually, an optimized factory with maxed out beacons, if compared with a factory without beacons but with the same output, does not have significantly higher energy demand. Usually, it has even less energy demand if used for items that need a small amount of time such as plates and circuits. This is due to the stacking of speed and productivity effects and due to the fact one beacon affects multiple assembling machines. A plate factory with a large field of alternating lines of beacons and furnaces is extremely energy and resource efficient.

Re: Nuclear power causes 38 GW power draw on system

Posted: Thu Sep 09, 2021 4:46 pm
by Silari
Tertius wrote: Thu Sep 09, 2021 1:55 pm
Silari wrote: Wed Sep 08, 2021 4:08 am Beacons are terrible, IIRC, and possibly the worst one. Especially since late game bases tend to have so many.
Usually, an optimized factory with maxed out beacons, if compared with a factory without beacons but with the same output, does not have significantly higher energy demand. Usually, it has even less energy demand if used for items that need a small amount of time such as plates and circuits. This is due to the stacking of speed and productivity effects and due to the fact one beacon affects multiple assembling machines. A plate factory with a large field of alternating lines of beacons and furnaces is extremely energy and resource efficient.
This conversation is about the increased energy draw of entities after a low power situation to refill the energy buffer, not about their utility.