Factorio Forums

With steam you can safely power you base and laser defence on the same grid since laser turrets gets priority over everything else and you can provide maximum power output day and night.

When powering your base with solar power you rely on buffering to survive the night.
If your base uses more power than you have scaled your solar + accumulator farm for then you prioritise your production over your defence by using up the energy before the natives attack. In the day you want to ensure that your accumulators are charging fast enough to reach 100% before the dusk and in the night you want to make sure that your charge level is high enough to survive until dawn.

What you want is your power system to provide no more than the highest average consumption your power system can handle when supplying more means that you would then eat up your future savings. This is actually very easy. If you make an independent grid that has exactly enough solar, accumulators to power a constant drain then you can just compare your main energy storage to the level of that ideal grid and see if you are overconsuming your buffer or not. If you have 2 consumer power grids of different priority levels then you can disconnect your low priority grid when you notice that you have consumed more than you should have.
In the picture the radar is on the low priority grid while your laser turret is on the high priority grid (direct connection to solar and accumulators). The radar consumes more power than 3 solar panels and 2 accumulators can provide as a game-day average and will drain your accumulators faster than you want and even prevent the accumulators from charging in the day. But when the charge level gets below the sunlight sensor system the radar is powered off which allows the accumulator to charge up as long as your high priority grid doesn't use it all up. Once the accumulator is charged enough the low priority grid is connected again. If the lasers aren't using your power then the radar will toggle on and off continously so that it gets on average no more than necessary to also give accumulators a steady charge up before dusk.

Of course if the solar grid was big enough to supply power to both grids then the low priority grid will be powered on all the time. Also the high priority grid can drain the system completely. There's no point in limiting the rate at which the high priority grid can draw power since if it uses any power then it definitly needs it now and saving it for the future is pointless.

In an actual system
you would probably have your entire base that isn't dedicated to defence on the low priority network and your lasers and roboports for repairs on the high priority network. If your solar farm is big enough to power your defence then it will never fail no matter how much you expand your production base before upgrading your solar farm. Since your perimeter grows linearly when your area that you wall in grows quadratically you are unlikely to ever run out of power for lasers once your base gets fairly big. If your base needs 10 times more power than your defence then you can expand your potential production and area required 100 times before upgrading your power system and still have enough power for lasers.

Also you might want to power the combinator by it's own personal power grid with 1 solar panel and 1 accumulator in case you are going to put Creative Mode/Energy void powerful energy deleteting power system testing entities on the low prio grid. It will at least be able to switch the low grid off then.

Nice . You could halve the size of the sunlight sensor by using 3 radars and something which consumes 150kW such as 5 small pumps on a water loop.

BlakeMW wrote:Nice . You could halve the size of the sunlight sensor by using 3 radars and something which consumes 150kW such as 5 small pumps on a water loop.

Yeah. But this is simpler q:
Can't place a blueprint with just enough liquid in it, but not too much so that it stops flow and so on. And you only need 1 sensor for your entire power grid so it should fit in somewhere anyways.
If you have enough solar to power your base and and have enough for all your laser defence and need this, then it's small and cheap to everything else you have I think.

Another idea would be just reduce everything in the sensor by half, with 3 radars. Then your base would power off to keep a bit higher than necessary accumulator charge if you want to be warned about the low power situation before it happens and keep a higher buffer for lasers in case your solar game day average production isn't enough for the worst attacks.

If a tiny (or big) underconsumption in the sensor is acceptable then you might be able to shrink it to 6 solar + 5 accus or similar with 1 radar or something less for extreme compression. A flickering toggle for the radar could approximate any consumtion and you might be able to build something good with with 1 solar, 1 accu and 1 radar!

Glad you like it.

Qon wrote:

BlakeMW wrote:Nice . You could halve the size of the sunlight sensor by using 3 radars and something which consumes 150kW such as 5 small pumps on a water loop.

Yeah. But this is simpler q:
Can't place a blueprint with just enough liquid in it, but not too much so that it stops flow and so on. And you only need 1 sensor for your entire power grid so it should fit in somewhere anyways.

True on not blueprintable, but a small pump loop (with no storage tank) is not all sensitive, as long as it's anywhere between almost completely full or almost completely empty it'll draw 150kW. A completely empty loop with only 3 water added will draw 150kW - a similar amount of slack in an almost full pipe will work.

Qon wrote:A flickering toggle for the radar could approximate any consumtion and you might be able to build something good with with 1 solar, 1 accu and 1 radar!

Challenge accepted!
Radar combinators:

Constant: R = 1
Decider: R < 100 (looped)
Power switch: R < 14

This runs the radar 13% of the time, consuming an average of 39 kW. The other 3 kW is eaten by combinators.

The output combinators multiply the accumulator level by 100/84, since it can never fully charge.

DaveMcW wrote:

Qon wrote:A flickering toggle for the radar could approximate any consumtion and you might be able to build something good with with 1 solar, 1 accu and 1 radar!

Challenge accepted! This runs the radar 13% of the time, consuming an average of 39 kW. The other 3 kW is eaten by combinators.

The output combinators multiply the accumulator level by 100/84, since it can never fully charge.

Fantastic! It's so tiny.

I guess you would place the radar before the panel in that case though.
With 2 solar panels and a flickering switch you could get something that is effectivly 1/0.84 solar panels, perfect for a single whole accumulator and which doesn't depend on build order. It would need more switches, combinators and one more solar panel though so not as compact. But I like having my systems stable enough that I can shortcut through my base with a FuPower tank or plasma shells and let my QonBots repair all damage. If something depends on build order then it's not automatically repairable. Same as why I avoid the liquid pump for energy draw.

But even better would be to, with more combinators, always completely fill the system to reset it. If it started draining before it hits 100% then it can enter a reset state where the radar is cut off until it's at 100% charge again. It would be fail-safe so your base might be turned off for half a day while it is resetting but then after that it would be in sync until you destroy it again!

Any takers? q:

With the single panel setup, you could add an extra constant + decider combinator that would inject R = -100 to the power switch while the charge level is above 84%. Then change the ratio to <20 / <150 for an average of 38 kW.

DaveMcW wrote:With the single panel setup, you could add an extra constant + decider combinator that would inject R = -100 to the power switch while the charge level is above 84%. Then change the ratio to <20 / <150 for an average of 38 kW.

Sounds like it would work perfectly even when blueprinted! Thanks
When Wube fixes the blueprint wiring bug that is, which affects all wiring dependent systems.
I might update OP with it later as an alternative and BP string if I decide to implement it.

If you want I can post simple setup that simply enables/disables steam depending on state of accus - it works with two thresholds - enables power between min and max accumulator charge.

orzelek wrote:If you want I can post simple setup that simply enables/disables steam depending on state of accus - it works with two thresholds - enables power between min and max accumulator charge.

Feel free. People who come here are probably more interested in reliable power production overall and not just the focus I have in OP.

Qon wrote:

orzelek wrote:If you want I can post simple setup that simply enables/disables steam depending on state of accus - it works with two thresholds - enables power between min and max accumulator charge.

Feel free. People who come here are probably more interested in reliable power production overall and not just the focus I have in OP.

Considering the fact that there are nice guides around I usually keep my simple and most likely to big contraptions to myself

Here is a design with 2 solar panels and 1 accumulator. It's a bit bigger, but also a bit easier to understand.

Looks good Dave! Thanks for sharing it.