determining shield recharge energy consumption

[ponkan]

So I'm speccing my brand-new Dostoevsky, which has 2 Microwave Lasers, 2000 MJ/s each, at 50% duty cycle means that average power consumption is 2000 MJ/s total, peak 4000 MJ/s. I buy some Quad Shields XIII, which has "Shield protection recharge speed set to: 388 MJ/s", multiplied by four facings for 1552 MJ/s. So (theoretically) total energy usage is 3552 MJ/s while dogfighting, and I buy a Reactor II which provides 4400 MJ/s -- should be enough, right? Realizing that actual peak usage is more while activating my weapons, I make sure that my capacitor provides at least 16000 MJ of stored energy, enough energy for 4 seconds of sustained firing, which is the beam stability figure -- I ended up with Capacitor II providing 20000 MJ. I thought I was good to go.

I thought the reactor rating seemed ridiculously low. Constant firing of my weapons showed that while the weapon was actually firing my capacitor levels fell, since the reactor didn't provide enough peak output. But capacitor levels went back to full while the weapons recycled, so it was fine. To fully test it out, I SPECced 'til my shields were depleted, then came out, only to see my capacitors drain in about two seconds, and take about 55 seconds -- I timed it -- to fully recharge the shields (at which point my capacitor energy started recharging). Buying a Reactor V with 6900 MJ/s took the recharge time down to 27 seconds. Clearly recharging from a state of depletion takes more energy than simply maintaining the shields, but how much? How do I determine that rate so that I can choose the right reactor?

BTW, I ran the numbers, and total energy usage while using Reactor II was 55s * 4400 MJ/s + 20000 MJ, or about 260 GJ (1000 MJ=1 GJ). Under Reactor V it was about 206 GJ, that's over 25% difference and way outside margin of error I think. So I try removing shield maintenance rate from reactor output, and things line up a little better:
55*(4400-388*4)+20000=176640
27*(6900-388*4)+20000=164396
The spread is 12244, about 7%, which is probably accounted for by measurement error.

At this point, I figure on trying one more thing out, calculating peak energy consumption by the time it takes to deplete my capacitors. Equation is (capacitor energy)/(depletion time)+(reactor recharge rage)
Reactor V: 20000/7.14+6900=9701
Reactor II: 20000/3.69+4400=9820
Since we've determined that shield maintenance is always present, that amount should be subtracted to determine recharge drain.
9701-388*4=8149, or 2037.25 per facing
Interestingly, 9700 is exactly 25 times 388

Until the advent of better data, I will therefore assume that recharge is 525% above maintenance rate, per shield facing.

I would love any information from the game's designers or anyone who's looked at the source and saw the actual equation(s) used by the game.

[loki1950]

Not sure on the equations actually used but the values for each ship are all defined in units.csv which is just a text file with commas separating each value it can be opened in any spreadsheet app M$ excel or OOo calc if you do any edits with either make sure that you save to the proper format or use cvs_editor v2 that you can find in this thread there are versions for windows and linux.

Enjoy the Choice

[ponkan]

Can you specify the heading for the data? I checked units.csv, and the only likely one I found was Shield_Recharge, which is specified in VSD/sec (seems to be about 5.4 MJ), and corresponds to the energy used for shield maintenance -- the same data field for quadshield13__upgrades is 72 VSD/sec, or 388 MJ/sec. This is column AO under OOo, I guess that'll be the 41st field if 1-indexed, 40th if 0-indexed.

[loki1950]

That sounds right with cvs_editor v2 it displays the column headings and when you select a cell it shows how the data in that cell is organized such as some can have multiple entries and therefore it's more human readable as the format does not lead it's self to easy reading but is very convenient codewise.

Enjoy the Choice