[15.37] Boiler Output/Belt Throughput Shortages

[Patric20878]

Hello,
While updating my steam engine design for a big update to my thread, Optimized Steam Engine Setup, I have found two possible bugs:

1) Boilers can't seem to actually sustain steam output for 2 steam engines at max load. Whether it's in my 60 boiler, 120 engine setup or a basic 1 boiler, 2 engine one, the results are the same - steam very slowly depletes until production starts fluctuating. In my map, I have tested this to take approximately 25 minutes to lose 600 steam in a boiler and 2 engines, which makes for about 1 steam lost every 2.5s. BlakeMW posted more about the technicalities of it in the thread, if the info is needed.

To reproduce this, let the red belt fill up with coal on my map, then flip the power switch. You will see the boilers slowly losing steam, followed by the 2 engines each they're connected to, over roughly a 25 minute period. Additional coal will be needed on the red belt before roughly the 20 minute mark as well (related to second issue in this report). At roughly 25 minutes, production will regularly dip a few MW, for each time engines are out of steam.

(Below confirmed as not a bug)
2) Red belts seem to be just short of having the throughput to supply 120 steam engines at full load. It'd appear that it supplies something like 119.5-119.8 engines or something. As I discovered this issue around the same time i discovered the boiler output shortage, it seems to me very likely that this could be a similar issue, with the recurring theme of x producing just short of a round, even integer. If this isn't a bug, but is unintentional, please consider the trivial 1-2% throughput buff it needs to supply exactly 60 boilers/120 steam engines at max load.

To reproduce this, let the red belt fill up with coal on my map, then flip the power switch. In roughly 20 minutes, the last couple steam engines will start fluctuating with coal outages. And from comparing their downtime with their uptime, it'd appear that the red belt is able to supply an estimate of 119.5-119.8 steam engines at max load.

As these two issues seem to be related, and are testable and reproducible in the same scenario I have provided below, I have listed both in the same thread, for ease of access. Please let me know if it's more preferable to keep these two in separate threads instead. Thanks for your investigation into this.

Edit: Please refer to BlakeMW's post on Mon Nov 20, 2017 5:59 pm for a save file containing an isolated setup that reproduces the non-1:2 boiler:engine steam ratio bug.

[Loewchen]

Provide a complete save file where 1) can be observed, 2) is not a bug.

[Patric20878]

Provide a complete save file where 1) can be observed, 2) is not a bug.

Scenario replaced with save file. Monitor boilers, then steam engines to see steam deplete.

Is 2) intentionally set to be just 0.2-0.5 engine's worth of coal lower throughput than 120? If not, what would this issue be categorized as and where would I post a thread on it?

[Patashu]

Is 2) intentionally set to be just 0.2-0.5 engine's worth of coal lower throughput than 120? If not, what would this issue be categorized as and where would I post a thread on it?

The impression I'm getting is that it's simply coincidence that the red belt's ability to sustain boilers being just shy of a nice round number is a coincidence, no more meaningful than any other numerology.

[Patric20878]

It's not just any round number though - yellow belts would be just short of the supplying the standard 30 boiler chain setup too.
Since mine has twice the length, red belts fall just short of twice the standard. And with how Blake guessed there may be some floating point imprecision or whatnot, the boiler/belt shortages seem like they could be related.

[Koub]

2) is sheer coïncidence. Belts have some throughput, boilers a consumption, and it just happen that one is very close to being multiple of the other.

[Patric20878]

Okay. I'll cross 2) off then. Maybe it fits more as a suggestion instead then to tweak it slightly so it is a clean multiple - the buff would be minimal.

Edit: Just additionally mathematically confirmed 2) to be not a bug, good.

[BlakeMW]

The boiler ratio thing is definitely a bug - it might be a stupidly tiny not worth fixing bug, but it's a bug because the raws say the Boiler provides 1800KW and the Steam Engines consumes exactly 900KW yet when both are at full load they are not matched in 1:2 ratio. Machines like furnaces have been fixed so they are consistent with their consumption numbers.

I've attached a save (made in sandbox scenario) of the simplest possible 1:2 scenario. You don't need to use accelerate time. Please mouseover the boiler, at save load it should have 53.5 steam in it, but it can be observed that every time a fuel is consumed the quantity of steam in the boiler falls by one. Originally I had thought it was a floating point issue, but it clearly isn't: the issue is just obscured when using long duration fuels like rocket fuel. Rather it seems to be exactly the "one tick of production lost on cycle end" issue.

steam loss.zip

(1.88 MiB) Downloaded 151 times

edit: Turns out there is already a bug report [0.15.13] Boiler is idle for a tick after craft cycle

[Elecen]

I have a similar but a bit more nightmare case. My boilers nuclear (but also with coal boilers earlier) seem to not be producing. They have wrong indicators, wrong numbers and if you check the Electricity network capacity - it should be at almost 1GW but looks like it's on 0.5GW (due to boilers issue) - and according to my reactors setup I could be having much more boilers.
Also, from time to time (no changes in setup or anything) some boilers come back to life. Especially after loading the game a few times.

I've attached two screenshots of the same game in different times - the exact same setup but different availability and the save game itself

screenshots

20171208000631_1.jpg (563.62 KiB) Viewed 5263 times

20171212113504_1.jpg (761.13 KiB) Viewed 5263 times

[Loewchen]

I have a similar but a bit more nightmare case. My boilers nuclear (but also with coal boilers earlier) seem to not be producing. They have wrong indicators, wrong numbers and if you check the Electricity network capacity - it should be at almost 1GW but looks like it's on 0.5GW (due to boilers issue) - and according to my reactors setup I could be having much more boilers.
Also, from time to time (no changes in setup or anything) some boilers come back to life. Especially after loading the game a few times.

This is not a bug, your heat exchangers are too far away from your reactors, you can get some improvement by adding a heat pipe to the west of your exchangers, but the setup does not take the heat gradient into account and therefore does not work properly.