Inspired by recent posts in the Changes Discussion Thread, I decided to investigate more carefully the question of how much maintenance storage should be used (if any). I took a ship design from my current game and systematically varied the balance between engineering spaces and maintenance storage bays, keeping the same overall tonnage. The ship is 32,000 tons with Magneto-plasma level tech, 18 months of deployment time, a Max Repair of 504 MSP, and a wide variety of systems including missiles, particle beams, turrets, sensors, and a small 500 ton hangar. I added engineering spaces one at a time and filled in the rest of the tonnage with maintenance. The results are as follows:
Engineering tonnage | Maint tonnage | Maint Life | MSP | IFR |
50 | 1106.5 | 0.69 | 11193 | 113.80% |
100 | 1051.5 | 1.16 | 10773 | 56.90% |
150 | 981.5 | 1.44 | 10202 | 37.90% |
200 | 921.5 | 1.7 | 9732 | 28.40% |
250 | 861.5 | 1.92 | 9262 | 22.80% |
300 | 801.5 | 2.08 | 8793 | 19.00% |
350 | 731.5 | 2.18 | 8224 | 16.30% |
400 | 671.5 | 2.27 | 7755 | 14.20% |
450 | 611.5 | 2.35 | 7286 | 12.60% |
500 | 551.5 | 2.4 | 6817 | 11.40% |
550 | 486.5 | 2.42 | 6299 | 10.30% |
600 | 426.5 | 2.44 | 5831 | 9.50% |
650 | 366.5 | 2.43 | 5364 | 8.80% |
700 | 306.5 | 2.41 | 4897 | 8.10% |
750 | 241.5 | 2.35 | 4380 | 7.60% |
800 | 176.5 | 2.27 | 3863 | 7.10% |
850 | 116.5 | 2.19 | 3397 | 6.70% |
900 | 56.5 | 2.08 | 2930 | 6.30% |
950 | 0 | 1.96 | 2499 | 6.00% |
These values will depend greatly on the details of your ship design -- its size, deployment time, Max Repair value, etc. But they suggest that if your only goal is to maximize deployment time, then an engineering:maintenance ratio of around 6:4 or 7:3 may be ideal, although using more engineering spaces doesn't hurt deployment time too much.
But of course maximizing deployment time is not the only goal one has when designing a ship. One issue to consider is the variance in your ship's maintenance life. While only using engineering spaces can give you a decent (although not optimal) nominal maintenance life, the rate at which you use MSP will have a very high variance: you might get lucky and have few failures, or you might get very unlucky, have several engine failures in a row, and run through all your MSP. Adding more MSP can help reduce this variance.
Another factor is the cost of MSP. While removing engineering spaces in favor of MSP storage can in some cases increase deployment time, over the lifetime of a ship the extra MSP usage can add up. The 2499 MSP of the all-engineering spaces design requires 249.9 Duranium and Gallicite and 124.95 Uridium, compared to 583.1 Duranium and Gallicite and 291.55 Uridium for a design with 12 engineering spaces (which gives the longest maintenance life) -- more than double the mineral usage. How much this ends up mattering will depend on the size of your fleet and how often it's out of port, but I think this can be substantial. (An engineering space requires 10 Duranium (plus more crew accommodations) compared to 1.5 Duranium and 1.5 Neutronium for a Maintenance Storage Bay of the same size, but I think over the lifetime of a ship these costs are much less than that of the MSP itself.)
What about weapon failures? The weapon failure rate is 1% as per the rules post. I added up the expected MSP usage from all my weapons assuming I empty the magazines and fire ten salvos with my beam weapons (this takes 150 seconds, the same amount of time it takes to empty the magazines). I got was the surprisingly low value of 153.7 MSP. This number will of course vary greatly depending on your ship design. For instance, if I replaced all missile launchers and magazines with particle beams this value would increase to 180.31 MSP. But this is still very small compared to the amount of MSP carried even by a ship with only engineering spaces, so it seems that weapon failures should not be a major concern. There are some caveats to this conclusion, though. One is again the issue of variance: while a 1% failure rate is very low, it means that the variance in MSP usage is fairly high. For instance, again using the all-particle beam design, there is a 12.4% probability of four or more weapon failures over the course of an engagement, requiring 379.6 MSP or more -- still not very much, but somewhat more substantial. Another factor to consider is that I've been running these numbers on a fairly large ship. A smaller ship will have fewer weapons but also fewer MSP, and I don't know how those will balance out.
There are probably lots of other factors that I've neglected here, and all of these numbers will depend greatly on your ship designs and on how they're used, but I've spent more than enough time on this already. I hope you find it useful! I know this is going to have a big impact on my ship designs going forwards.