Author Topic: Ship Maintenance Theory  (Read 7648 times)

0 Members and 1 Guest are viewing this topic.

Offline Alsadius (OP)

  • Lieutenant
  • *******
  • Posts: 176
  • Thanked: 87 times
Ship Maintenance Theory
« on: April 23, 2020, 05:40:44 PM »
Per a discussion on Discord, I've done a deep dive into maintenance cost usage, and come up with some interesting things. (At least, they're interesting if you're as big of a game mechanics geek as me). I'll also cover the basics, in case newbies get linked here. Points where I'm still uncertain are in italics. I've tried to confirm things where I can, but I may have missed some details, so feel free to correct me in the comments if needed.

Here's Steve's post, which I used as a reference for a lot of this.

Military ships use maintenance supply points (MSP), but the maintenance formulas are different if a ship is at a maintenance facility, in space, or in overhaul. Commercial ships do not use MSP in normal circumstances.

Ships at Base
If the ship is at a base(i.e., a planet or ship with maintenance facilities), and the base is large enough to fully maintain the ships stationed there, the formula is simple. The ship uses MSP equal to [cost in BP*25%] while it's sitting idle. If it's in overhaul, the ship will use 4x this amount, or [cost in BP*100%].

Each MSP costs 0.25 BP (0.1 Duranium, 0.1 Gallicite, and 0.05 Uridium). As a result, this means that a military ship will consume 25%*0.25 = 6.25% of its BP cost per year in maintenance costs, even if it never leaves base. Note that this is the same for all military ships - your engineering spaces do not matter at all. A year in overhaul costs 25% as much as a new ship.

Ships in Space
If a military ship leaves base, then it stops following the above rule. Instead, it follows a different set of rules, best summarized in the class design window.

Here's the maintenance line for one of my light carriers. For reference, it's a 9,995 ton ship, cost 1,136 BP, with two engineering spaces and three small maintenance storage bays.

Quote
Maint Life 0.92 Years     MSP 382    AFR 400%    IFR 5.5%    1YR 417    5YR 6 262    Max Repair 250.00 MSP

In order, these mean:
  • Maint Life: On average, the ship will suffer enough failures in 0.92 years to completely drain its maintenance supplies, and be left unable to fix modules as they break.
  • MSP: The ship carries 382 MSP.
  • AFR: The ship has an Annual Failure Rate of 400%. In other words, in an average one-year deployment, the ship will suffer four item failure events.
  • IFR: The ship has an Interval Failure Rate of 5.5%. In other words, in an average five-day interval, the ship has a 5.5% chance of an item failure event. Since there are 73 intervals per year (365 days / 5 days per interval), the AFR is always 73 times higher than the IFR.
  • 1YR: In a one-year deployment, the ship will need an average of 417 MSP to repair failures.
  • 5YR: In a five-year deployment, the ship will need an average of 6,262 MSP to repair failures. Note that this is a lot more than 5x the 1YR value, which we'll get to below
  • Max Repair: This is the cost of the single most expensive item that can break on the ship.

So, if this ship is in base, it will cost 1,136 BP*25% = 284 MSP per year. If it is in space, it will instead cost an average of 417 MSP for the first year. And it's important to keep in mind that all of these are averages. The failure chance is calculated once per interval, and then if you get a failure, the broken item is chosen at random. With bad luck, a ship that doesn't have enough MSP to do its max repair can fail in the very first interval it's in space, even if the Maint Life value is several years.

However, there's another very important angle on deployments. As the ship has been deployed longer, the failure rate will increase over time. In the second year, the failure rate is twice as high. In the third year, it's 3x as high, and so on. So for that ship, we have the following MSP costs:
  • Year 1: 417 MSP
  • Year 2: 417*2 = 834 MSP
  • Year 3: 417*3 = 1,251 MSP
  • Year 4: 417*4 = 1,668 MSP
  • Year 5: 417*5 = 2,085 MSP
  • TOTAL: 417+834+1,251+1,668+2,085 = 6,255 MSP (This differs slightly from the 6,262 above because of rounding errors in my math)

So even if you don't mind the 417 MSP cost for deploying in the first year, it'll rapidly increase over time. This is why long deployments are difficult for military ships. This is also why overhaul is a thing - simply returning to base won't reset the maintenance clock, it'll just stop it from growing. To get the ship back down to the nice low year-1 failure rates, the ship should be overhauled periodically.

Technical note: I'm not 100% sure that the failure rate is based strictly on years the way I've implied - it's possible that months matter too, and a failure less likely in month 1 than it is in month 11. However, the formulas on the ship design screen seem to imply it's flat throughout each year.)

If a ship is at a base, but the base doesn't have enough maintenance facilities to support all the ships there, maintenance uses a weighted average of the in-base and deployment systems. See Steve's post for details.

Failure Rates
The listed failure rate for a military ship is determined by two things - total tonnage and engineering space tonnage.

If a ship has no engineering spaces, the AFR is equal to [20%*tonnage]. So a 10,000 ton ship with no engineering spaces will have an AFR of [20%*10,000] = 2,000%. Obviously, that's not ideal. This is why we add engineering spaces. If a ship has any engineering spaces, the AFR is equal to [0.04%/engineering tonnage percentage*total tonnage]. The engineering tonnage percentage is just the percentage of the total tonnage of the ship that's made up of engineering bays. For my ship, with 100 tons of engineering and 9,995 total tons, this is a hair over 1%. So [0.04%/1.0005*9,995]=399.6% AFR.

Note that it's technically possible for the second formula to be worse than the first. If the ship is less than 0.2% engineering, it's better to have none at all. Of course, a ship of any size with less than 0.2% engineering is a maintenance nightmare to begin with, but if you're going to do that, pull the engineering spaces off entirely.

If you increase the engineering tonnage percentage, you get directly proportional reductions in IFR/AFR and 1YR/5YR MSP usage. Double the engineering spaces(while holding ship size constant), and you halve the failure rates and MSP usage.

Maintenance storage bays do not play any role in this calculation. Instead, they simply give you more MSP to play with. So if you get a 4-month life on your ship with engineering spaces, and you add a maint bay to double your MSP in the ship (again, holding size constant), you'll have an eight-month life. The same is true if you burn through all your MSP and get resupplied from a supply ship - supply ships basically act like virtual maintenance supply bays for this purpose.

Also, listed failure rates aren't the full story. Ship crew and commander skills play a role here. The crew grade bonus reduces the failure chance, as does the Engineering skill of its command team. Half the commander's skill, plus the Chief Engineer's skill, is applied to AFR/IFR values. As well, the bonuses from fleet commands can apply to ships in the appropriate radius (25% from Naval, Patrol, or Survey, and 10% from General). This means that your good ships can last longer than they look like they'd be able to, especially if they have a good command team.

Repairing Damage
All of the above discussion was about repairing normal damage due to wear and tear. Combat damage follows different rules.

(Edited damage control info due to comments below and further research)

When an item is damaged in combat, the damaged ship doesn't instantly repair the way it does for normal failure. Instead, the ship's Damage Control Rating (DCR) is used. A ship gains 1 DCR per engineering space, and 10 DCR per Damage Control module. The DCR is the average cost of modules that the ship can repair in 1000 seconds (Link). When a ship uses DCR to repair, the repair cost is twice as high as the item's cost in BP. Note that armour cannot be repaired with DCR.

Any non-missile weapons have a failure chance of 1% each time they're fired. This consumes MSP equal to their BP cost, but doesn't require DCR.

For ships expecting to fire non-missile weapons and/or take combat damage, you will often want to have extra MSP around to make repairs in combat. If you do that, remember not to double-count your MSP. If you have 1000 MSP on your ship, but you expect 500 of it to be used for combat repairs, then that "10-month" maintenance life is really only five months, because half your supplies are earmarked elsewhere. The game will not do this calculation for you, so keep it in mind. It's often worth keeping a maintenance life that's a little too long for your deployment time, if your ship is in this category.

Optimization Notes
While the average failure rate doesn't depend on the ship's design, the average failure cost does. Not all items in a ship can fail (e.g., there are no armour failures), and since repair cost is based on item cost, the costs of your breakable items will affect the MSP spent on repairs. If you take two engines of the same size and tech, the one with a higher power boost will be more expensive to build. Putting that on a ship will increase the annual MSP cost, because it fails equally often and costs more when it does.

On that note, remember that the failure happens at the ship level(the IFR calculation every interval), and only then is an item chosen. So an equal weight of small sensors(for demonstration, say 100x10 tons @ 1 BP each) might cost 2x more than a single engine(1x1000 tons @ 50 BP), but because the same failure hits a ten-ton sensor instead of a thousand-ton engine, the engine will cost 50x more to repair. We all know that large engines are more fuel-efficient, but large engines are less MSP-efficient when a ship is deployed.

For ships with extremely short deployment times, like fighters, you might launch, fight, and return without an interval ticking. No interval = no failure chance. Fighters don't have a high failure rate in the first place, because they're so small, but even still you can get away with cutting an engineering space a bit more readily than you might think.

Note that it's possible for deployment to be cheaper than sitting at a base, especially if you keep your deployments short and your ships well-engineered. Don't feel like sitting at base is the only option to save MSP. However, if you're going to max-min like that, don't forget overhaul costs.

The 6.25%/year cost to maintain a ship is meaningful - every 16 years, you have to pay the cost of the ship again, just to keep it in base. This is especially a concern if you're facing Gallicite/Duranium/Uridium shortages, since those are what MSP cost to build. For very obsolete ships, it might be better to scrap it just to save on the MSP. If a ship is coming back from a long deployment, this goes double - if it'll spend a year in overhaul, you'll be spending a quarter of its new-build cost to get it back in fighting shape. Consider whether that's a good investment, or whether refit/scrapping would be better.
« Last Edit: April 24, 2020, 05:34:50 AM by Alsadius »
 

Offline Bremen

  • Commodore
  • **********
  • B
  • Posts: 743
  • Thanked: 150 times
Re: Ship Maintenance Theory
« Reply #1 on: April 23, 2020, 06:19:21 PM »
The 6.25%/year cost to maintain a ship is meaningful - every 16 years, you have to pay the cost of the ship again, just to keep it in base. This is especially a concern if you're facing Gallicite/Duranium/Uridium shortages, since those are what MSP cost to build. For very obsolete ships, it might be better to scrap it just to save on the MSP. If a ship is coming back from a long deployment, this goes double - if it'll spend a year in overhaul, you'll be spending a quarter of its new-build cost to get it back in fighting shape. Consider whether that's a good investment, or whether refit/scrapping would be better.

This also means that if your ship has more than 16 years maintenance life you can cut costs by just leaving it on deployment until it blows up from a critical maintenance failure :P. I've done that with very long deployment FACs as jump point guards before.
 

Offline Jorgen_CAB

  • Admiral of the Fleet
  • ***********
  • J
  • Posts: 2822
  • Thanked: 673 times
Re: Ship Maintenance Theory
« Reply #2 on: April 23, 2020, 06:31:57 PM »
Pretty good overview... ships actually have become a bit more expensive since VB6 version. I think the cost for paying total ship cost there was 20 years, but I could be wrong though.

It can be quite valuable to keep your maintenance life value to around 2-3 times the ships deployment time. Every time you retire the ship to port for resting the crew you also overhaul the ship. Also, the more engineering sections you have you also have a higher chance that you don't get any failures. If the ship have maintenance life of three times the deployment time there is a very good chance you get zero failures.

You also should consider not using maintenance storage on warships. You are way better of adding more maintenance facilities to reduce the chance of failure instead. Make sure to have supply ships close at hand instead.

I usually keep my maintenance life of ships in the range of 1.5-2.5 years for most regular warships that typically have between 6-12 month of deployment time. It is a fairly good trade off between space and economy in the long run. Engineering spaces will generally keep maintenance costs down. You also should consider and engineering station on larger ships as that will keep maintenance failure even further down, especially with a good officer in charge.
« Last Edit: April 23, 2020, 06:34:03 PM by Jorgen_CAB »
 

Offline Alsadius (OP)

  • Lieutenant
  • *******
  • Posts: 176
  • Thanked: 87 times
Re: Ship Maintenance Theory
« Reply #3 on: April 23, 2020, 06:52:14 PM »
Pretty good overview... ships actually have become a bit more expensive since VB6 version. I think the cost for paying total ship cost there was 20 years, but I could be wrong though.

It can be quite valuable to keep your maintenance life value to around 2-3 times the ships deployment time. Every time you retire the ship to port for resting the crew you also overhaul the ship. Also, the more engineering sections you have you also have a higher chance that you don't get any failures. If the ship have maintenance life of three times the deployment time there is a very good chance you get zero failures.

You also should consider not using maintenance storage on warships. You are way better of adding more maintenance facilities to reduce the chance of failure instead. Make sure to have supply ships close at hand instead.

I usually keep my maintenance life of ships in the range of 1.5-2.5 years for most regular warships that typically have between 6-12 month of deployment time. It is a fairly good trade off between space and economy in the long run. Engineering spaces will generally keep maintenance costs down. You also should consider and engineering station on larger ships as that will keep maintenance failure even further down, especially with a good officer in charge.

I like maint bays on warships, to a certain limited extent. I don't like relying on my (slower, undefended) supply ships keeping pace unless I have to, and spending 50 tons to get my MSP up is often a good investment. Doubly so if the ship needs to refill MSP for parasites, or if it has beams that can break in combat. Maint storage is often a better way to keep your beams firing than engineering spaces are, because your supply ships should really not be entering beam range, and the warships can't easily pull back to refill on MSP.

Offline Migi

  • Captain
  • **********
  • Posts: 465
  • Thanked: 172 times
Re: Ship Maintenance Theory
« Reply #4 on: April 23, 2020, 08:34:57 PM »
Technical note: I'm not 100% sure that the failure rate is based strictly on years the way I've implied - it's possible that months matter too, and a failure less likely in month 1 than it is in month 11. However, the formulas on the ship design screen seem to imply it's flat throughout each year.)
I suspect that maintenance failure rate is based on the ship maintenance clock, so it would increase every construction cycle rather than having big jumps every year.

When an item is damaged in combat, the damaged ship cannot self-repair unless it has a Damage Control module. Even if it does, the repair cost is twice as high as the item's cost in BP.

When you say damage control module I assume you mean damage control rating, which is provided in small amounts by engineering spaces.
 

Offline Alsadius (OP)

  • Lieutenant
  • *******
  • Posts: 176
  • Thanked: 87 times
Re: Ship Maintenance Theory
« Reply #5 on: April 23, 2020, 08:44:29 PM »
I suspect that maintenance failure rate is based on the ship maintenance clock, so it would increase every construction cycle rather than having big jumps every year.

I wouldn't be surprised if that's the case. But if it is, Steve's formulas in the class design window are giving us bad data.

When you say damage control module I assume you mean damage control rating, which is provided in small amounts by engineering spaces.

This one I should have flagged as uncertain. Testing, it seems like you're right - engineering spaces give 1 damage control rating each, and a damage control gives 10(albeit, on a module 3x the size). I don't understand how DCR works, though, so I suspect I need to amend that section.

Offline Father Tim

  • Vice Admiral
  • **********
  • Posts: 2162
  • Thanked: 531 times
Re: Ship Maintenance Theory
« Reply #6 on: April 23, 2020, 08:45:15 PM »
Technical note: I'm not 100% sure that the failure rate is based strictly on years the way I've implied - it's possible that months matter too, and a failure less likely in month 1 than it is in month 11. However, the formulas on the ship design screen seem to imply it's flat throughout each year.)
I suspect that maintenance failure rate is based on the ship maintenance clock, so it would increase every construction cycle rather than having big jumps every year.

When an item is damaged in combat, the damaged ship cannot self-repair unless it has a Damage Control module. Even if it does, the repair cost is twice as high as the item's cost in BP.

When you say damage control module I assume you mean damage control rating, which is provided in small amounts by engineering spaces.


Migi is correct on both points.
 

Offline Ri0Rdian

  • Warrant Officer, Class 1
  • *****
  • Posts: 93
  • Thanked: 15 times
  • Discord Username: Ri0Rdian#3639
Re: Ship Maintenance Theory
« Reply #7 on: April 23, 2020, 09:12:35 PM »
Thank you for this post. I am a big theorycrafter and love to know ins and outs of games I play so I can do my best (or at least not suck  ;D ). This certainly helps a lot. As I got older I lose paitence to do this stuff myself so such detailed information is much appreciated!
 
The following users thanked this post: Alsadius

Offline skoormit

  • Commodore
  • **********
  • Posts: 764
  • Thanked: 310 times
Re: Ship Maintenance Theory
« Reply #8 on: April 23, 2020, 10:31:28 PM »
...
If a ship has any engineering spaces, the AFR is equal to [0.04%/engineering tonnage percentage*total tonnage].
...

This is the same as:  (Total HS ^ 2) / (Eng Tons)

(Keeping in mind that 1 HS = 50 tons.)

In your example, 9995Tons = 199.9HS.
So: 199.9^2 / 100 = 399.6001
 

Offline TheDeadlyShoe

  • Vice Admiral
  • **********
  • Posts: 1264
  • Thanked: 58 times
  • Dance Commander
Re: Ship Maintenance Theory
« Reply #9 on: April 24, 2020, 12:17:53 AM »
I tend to overhaul everytime i do shoreleave just for useability.   Though its an interesting thought to do 20 year lifetime ships and just never overhaul. If you wanna keep the ship just refit it to newer tech and overhaul the rest of the time off. xD

I kinda wish Damage Control was needed for actual repair, it would give a point to that component.. ;p
 

Offline Jorgen_CAB

  • Admiral of the Fleet
  • ***********
  • J
  • Posts: 2822
  • Thanked: 673 times
Re: Ship Maintenance Theory
« Reply #10 on: April 24, 2020, 02:39:10 AM »
Pretty good overview... ships actually have become a bit more expensive since VB6 version. I think the cost for paying total ship cost there was 20 years, but I could be wrong though.

It can be quite valuable to keep your maintenance life value to around 2-3 times the ships deployment time. Every time you retire the ship to port for resting the crew you also overhaul the ship. Also, the more engineering sections you have you also have a higher chance that you don't get any failures. If the ship have maintenance life of three times the deployment time there is a very good chance you get zero failures.

You also should consider not using maintenance storage on warships. You are way better of adding more maintenance facilities to reduce the chance of failure instead. Make sure to have supply ships close at hand instead.

I usually keep my maintenance life of ships in the range of 1.5-2.5 years for most regular warships that typically have between 6-12 month of deployment time. It is a fairly good trade off between space and economy in the long run. Engineering spaces will generally keep maintenance costs down. You also should consider and engineering station on larger ships as that will keep maintenance failure even further down, especially with a good officer in charge.

I like maint bays on warships, to a certain limited extent. I don't like relying on my (slower, undefended) supply ships keeping pace unless I have to, and spending 50 tons to get my MSP up is often a good investment. Doubly so if the ship needs to refill MSP for parasites, or if it has beams that can break in combat. Maint storage is often a better way to keep your beams firing than engineering spaces are, because your supply ships should really not be entering beam range, and the warships can't easily pull back to refill on MSP.

Maintenance bays do have some uses but overall they are not needed all that much. Engineering sections will both reduce the need for MSP and increase the MSP at the same time.

If you build a dedicated beam ship you "might" need some extra MSP, but even then it should be rare to run out of MSP in most cases. I think it mostly is for planetary bombardment purposes that you will run out of MSP that way. If the beam ship has a health overall maintenance life cycle it is less likely to drain its MSP on their way to the fight and have more MSP from the Engineering sections too.

You also can have relatively fast supply ships too, but I never really had much problem with my supply train anyway in Aurora when I needed them.
 

Offline Eretzu

  • Warrant Officer, Class 2
  • ****
  • E
  • Posts: 52
  • Thanked: 22 times
Re: Ship Maintenance Theory
« Reply #11 on: April 24, 2020, 02:44:51 AM »
Talking about shore leave.  Is there a separate order for that or is follow order needed? (for example survey ships shore leave)
 

Offline SpikeTheHobbitMage

  • Bug Moderators
  • Commodore
  • ***
  • S
  • Posts: 670
  • Thanked: 159 times
Re: Ship Maintenance Theory
« Reply #12 on: April 24, 2020, 04:52:07 AM »
Talking about shore leave.  Is there a separate order for that or is follow order needed? (for example survey ships shore leave)
There is no separate order, but overhaul usually takes longer and counts as leave time.
 

Offline Alsadius (OP)

  • Lieutenant
  • *******
  • Posts: 176
  • Thanked: 87 times
Re: Ship Maintenance Theory
« Reply #13 on: April 24, 2020, 05:41:01 AM »
Thanks to the comments here, I did a bit of research and found a description of the damage control rules. I've edited that section.

...
If a ship has any engineering spaces, the AFR is equal to [0.04%/engineering tonnage percentage*total tonnage].
...

This is the same as:  (Total HS ^ 2) / (Eng Tons)

(Keeping in mind that 1 HS = 50 tons.)

In your example, 9995Tons = 199.9HS.
So: 199.9^2 / 100 = 399.6001

Yup. There's a bunch of ways to show the math. I figured the formula that I used was the most useful in normal circumstances, but if there's something specific you want to look at, you can and should play with the math here.

Offline skoormit

  • Commodore
  • **********
  • Posts: 764
  • Thanked: 310 times
Re: Ship Maintenance Theory
« Reply #14 on: April 24, 2020, 05:55:41 AM »
Quote from: Alsadius
...
If a ship has any engineering spaces, the AFR is equal to [0.04%/engineering tonnage percentage*total tonnage].
...

This is the same as:  (Total HS ^ 2) / (Eng Tons)

(Keeping in mind that 1 HS = 50 tons.)

In your example, 9995Tons = 199.9HS.
So: 199.9^2 / 100 = 399.6001

Yup. There's a bunch of ways to show the math. I figured the formula that I used was the most useful in normal circumstances, but if there's something specific you want to look at, you can and should play with the math here.

The major takeaway for me is that failure rate is proportional to the square of ship size, but inversely proportional to the (non-squared) engineering tonnage.
Meaning that if you double the ship size, you have to quadruple the engineering tonnage to keep the same AFR.