Balance electric charge?

[Magzimum]

First post! Hello everybody.

What methods do you guys use to decide how many solar panels (or fuel cells) and batteries you need on your craft? I have been playing for a few weeks, and it appears that my decisions on the number of solar panels on my crafts are determined more by the available space on the hull than on its needs in terms of consumption of charge.

If I will be doing stuff in the dark, I stack more batteries onto the craft, but again, no actual method to choose how many... the number is generally chosen by the space in the service bay.

The only actual thought (trick) that I use is to always disable one battery (I typically use 8 Z-100's of which I disable one, rather than 1 Z-k1), so that I can activate that to have some charge in case of an emergency (like when I forgot to extend the solar panels again, and only find out when I arrive at my destination).

[bewing]

Welcome to the forums!

The Z100's are a much better deal for cost of course -- just so long as they are inside a bay, so they don't get exposed to drag and heat. For a krewed ship at the beginning of the game, your main electric use (when you get to that point) is going to be transmitting science back to KSC. A big science transmission can use up 330 zaps. And you often have to do three or four transmissions at a time when you get to a new biome. So 800 zaps worth of battery is a minimum, but a nice minimum. Maybe you have to wait a few minutes for the batteries to recharge before you can do the final transmission.

When you get to the point where you don't really want to have service bays on your rockets anymore, then the Z1k's start to be nice. Another point in their favor is when part counts start to matter in the VAB. If you have a 30 part limit, and you need some electrical storage, then one big battery is the way to go.

The Z200's are sometimes nice on spaceplanes that also have a small reaction wheel, for reentry. Holding an AoA during reentry often takes a little over 100 zaps. So storing 200 is a convenient amount

The two real scenarios where you need huge batteries are drilling for ore/converting ore to fuel -- and vessels with multiple ion drives. When you are drilling and converting, 5000 zaps of battery is perfectly reasonable. A multi-ion drive ship with 10000 zaps of storage is also perfectly reasonable.

The main issue with solar is more one of "if you do something to drain your batteries way down, what's an acceptable time to charge them back up again?" And usually that time is fairly long. 5 minutes? So you often don't need much in the way of solar. A bigger consideration with solar is arranging enough of them so that at least one is always in the light (when there is light). Another is leaving enough room so you can dock without breaking your panels off.

But again -- drilling/converting and ion drives require a handful of gigantors. You may need 50 zaps per second of continuous power for drilling & converting.

[radonek]

I learned the hard way to always put a few ox-stats around to ensure some minimal power. Similar weight and less fuss then disabled battery. Rest depends on mission profile - while manned kerbaled Mun landing can be done on capsule batteries, sending gravioli scan from dark side of orbit will obviously need more. It's not like we are talking rocket science… oh wait. Ok, here's one: do not even try to have "enough" battery power for ion drive. If you find yourself doing (nontrivial) ion burn in the dark, you are just doing it wrong. Spending few grams of xenon to alter orbit is better then hauling around ton of batteries for dark side burn, even if that burn is actually more optimal.
 

[Magzimum]

Interesting to read that you do consider the electricity an issue of "instantaneous" consumption of charge (zaps), and charging time. I never looked at it that way. I looked at it from a point of balancing production and consumption... And yes, science transmissions are a huge spike in consumption which drain the reserves, and so is using the SAS on the dark side of any planet/moon... but regular SAS use is a more-or-less constant consumption which I try to balance. 

I appreciate the answers... made me look at things in a different way, which is always helpful.

Btw, I try to minimize the science transmissions from craft without a science lab, as they yield lower science points. And I find that the science labs are pretty much useless if you don't have a 5-star scientist, which I don't have (yet). It takes a good part of eternity to get any science out of it, and in game-play-time it seems quicker to just fly home. So, stubborn as I am, I rather fly the science home than to transmit it. (As for one-way probes, I learned about the amount of charge needed for a large data set (e.g. atmosphere composition) the hard way... by landing on Eve and not having enough batteries on board).

[Snark]

Here's a helpful way to think about it.  There are basically two kinds of ships in KSP:

• High-power
• Low-power

"High-power" ships are ones that need LOTS of electricity.  Most ships don't; there are only a few parts in KSP that really suck power.  Here are the in-game things that use lots and lots of electricity over extended periods of time:

• Mining drills
• ISRU refineries
• Ion drives
• Science labs doing research
• Rover wheels, when used for extended periods of long-distance travel

...note that all of the above use the high power only when actively turned on.

If you've got a ship with any of the above, then you need to balance your electric power, yes.  The way you do that is to add up the power requirements for the parts (you can get that from the info tab in the VAB, or off the KSP wiki) to figure out what your total EC/second requirements are, and then put enough solar panels on to meet the demand.  You can get the solar power generation numbers the same way-- just be sure to add some safety margin (the quoted numbers are for when the panels are perfectly face-on to the sun, which you generally won't be), and take into account that solar power generation varies with distance from the sun.  Fuel cells are a nice, lightweight, easy way to generate scads of electricity, but they consume fuel, so they're mainly useful for mining/refining ships where you just use part of the fuel output to generate the electricity.

However, most of your ships (i.e. ships without any of the above) are low-power.  Typically the only permanent electricity drain they'll have is for a probe core, and that's generally pretty minimal; even a single OX-STAT panel can keep up with probe core demand, unless you're very far away from the sun (e.g. Jool).  There are other things that use power, yes, such as reaction wheels, but they only do it briefly and only when in active use, so you generally don't need to worry about that.  So for most ships, you don't really need to worry about balancing electric power generation at all-- a single pair of the small folding panels, or a single quartet of OX-STATs, will be good enough for just about everything.

The one thing that I haven't mentioned in the above is transmitting science results.  That uses a lot of electricity, yes.  But it's a single hit per transmission, so there what you need is not generation capacity, but battery capacity.  A good rule of thumb I like to use there is:  have at least 400 EC of battery storage in the early game (when I only have the low-end science instruments), and at least 1000 EC in the later game (when I have the accelerometer and gravimeter).  A single Z-1K battery does the job for most ships, and is a convenient form factor that fits nicely on the ship without exposure to drag.

[bewing]

1 hour ago, Magzimum said:

Btw, I try to minimize the science transmissions from craft without a science lab, as they yield lower science points.

That is a slight misunderstanding actually. If you take (say) a gravity reading and bring it back to KSC, let's say you get 50 points for that. Now let's say you take a reading, transmit it, take a second reading, and bring that one home. You will get 20 points for the transmitted data immediately, then 40 more when you get the second copy back to KSC. Total = 60. You actually get a little more total, because you got credit for two copies of the experiment, rather than just one. These numbers are just speculative in this case -- but you should verify that I'm wrong before you abandon the thought of transmitting experiments.

And I have to disagree with radonek about ion engines. Burns are always of limited length. So if you have a lot of batteries, you can eliminate one gigantor -- so the mass thing balances out.

[Zhetaan]

@Magzimum:

I see that you've already figured out the Oldest Trick in the Book^TM when it comes to handling electric charge.  Good for you!  Now let me show you how to design a mission plan rather than build a ship and go.

This is a long post, but I wanted to cover everything because my approach is a bit different from what others have said so far.  I tend to divide the power requirements into two parts:  one covers the constant-running equipment and the other covers the single-use power spikes.  I look at the first one and balance it according to duration, and the second according to consumption.

Power systems that work for you will change with the mission and the ship design:  you know this.  A fully equipped ion probe will need something different from a once-and-done orbiter.  The rules for power parts are not too difficult, but there are a couple of surprises.  My method, when I want to do more than throw batteries and panels at it until it stops complaining, is this:

Add up the power drain rates of everything that will run continuously (watch the units; some parts are listed in EC/minute, others in EC/second).  I find it easier to convert everything to EC/second because the solar panels are rated that way.  This value is your baseline and it includes things such as probe cores, but it also includes drills, ISRU, and science labs--anything that will come on and stay on for a long time.  If it has to run longer than an orbit (or on the surface, run overnight), then it's part of the baseline for me.

Next, find your orbital darkness time.  The formula is not easy, so here: have a calculator.  Multiply your darkness time by the power drain to get some number in EC:  this the minimum amount of battery you need to cover the dark part of the orbit.  If you're landing instead, then use the wiki or the in-game Knowledge Base (find it in the Tracking Station) to find the rotation period of whatever you're landing on and use that to figure the rest.  If you want to add a safety margin, this is when to do so.  In your case, this is when to add the battery that you lock.

Next, you'll want to figure the power generation you need to charge the batteries in one orbit.  For that, you need to know (roughly) your orbital period.  You can figure it out by going to map mode, finding the difference between time to periapsis and time to apoapsis, and doubling that value.  Subtract orbital darkness time from orbital period to get orbital light time.  Divide the battery EC you figured out earlier by the orbital light time to get some number in EC/second; this is the charge rate you need (if you're using solar panels).

Add the charge rate to the baseline power drain you got earlier:  this total is the absolute minimum required amount of power generator.  You need to bring this much, no matter what, or else you will run out of power.

To figure supplemental power, it's actually simpler.  You just figure out how much power it will cost to run the equipment when it's on, add in the battery capacity, and throw on enough generators to recharge the batteries in a convenient amount of time.  It doesn't have to be all in one orbit, because it isn't essential for control of the ship:  you instead make certain the batteries are full when you want to run the equipment, and add enough generators that the batteries will be full when you want to run the equipment again.

An example of a supplemental system is the antenna for science transmission; this costs a varying amount of power for each experiment based on the amount of data and the type of antenna, but expect to spend about a thousand EC for a full set of experiments.  Experiments have a long shelf life, so you can afford to wait for the charge to build back up before transmitting more results.

 

Snark is absolutely correct that for most purposes, one solar panel will do for all the power generation you need:  battery capacity is the important thing, and even then, a single Z-1k will cover most of your needs.  I include the more complicated parts of it because 1) you asked, and 2) I don't know whether you run mods, but if you do, a lot of them have power-hungry devices that run constantly, and they can make power system design a much more thorny problem:  Life support usually requires EC for recyclers and some more for your Kerbals, SCANSat needs it for its scanners, there are a few electric propellers out there (it's annoying, but not life-threatening, to run out of power on an ion probe--not so for an electric aeroplane!), RemoteTech needs it for its antennas, and there are many more.  Knowing how to plan for those mods will let you have more fun when you use them.

 

For generators, keep in mind that solar panels have a few caveats to them:  they only work at their rated value when they are at Kerbin, cool, out of the atmosphere, and oriented correctly.  Distance, heat, and air all interfere with panels' base production.  Distance should be obvious:  the further you are from the light source, the less light falls on you.  Heat reduces the internal efficiency of the panel, but it usually is a consequence of the environment and there is little to do about it.  Atmospheric effects are simply the scattering and absorption of solar energy by the air.  This is modelled pretty well; you get much less power near sunrise than at noon because at sunrise, the light has to go through more air.  For this reason, you may not want to use solar panels on the surface of Eve--Eve's atmosphere is both hot and thick.  But solar panels are great in Eve's orbit; they work nearly twice as well as at Kerbin.  Orientation is tricky with static panels; one solution is to put at least one facing each of the six directions from your rocket.  The tracking panels are nice, but the funny thing about them is that if they ever do run out of power, they don't track.  That can be a problem if you run out of power behind a moon and have your panels facing the wrong way when you see the sun again.  Another thing to keep in mind that rotation in one axis only lets the panel face in directions perpendicular to that axis--in other words, if you point the axis of rotation at the sun, your panel will never receive power no matter how it turns.  This happens to me frequently, so I always add at least one solar panel at right angles to the rest.  It looks odd, but it works.

Fuel cells work well, but they obviously only work for so long as you have fuel for them.  One key advantage to them is that they produce power at a constant rate, so it's a simple calculation to determine how much fuel you need to bring.

RTGs work well, also, especially since they don't require fuel, but they produce less power for the mass.  Each one takes .75 EC/second off the drain rate; but they generally are not suited to be a primary power supply for any except the smallest science probes.  Their advantage comes from the fact that they provide power all the time without your needing to do anything about it.

 

Anyway, sorry for the wall of text, but that's how I do it.  It may be overthinking, but I've never had power problems after doing it this way.

[Snark]

38 minutes ago, bewing said:

If you take (say) a gravity reading and bring it back to KSC, let's say you get 50 points for that. Now let's say you take a reading, transmit it, take a second reading, and bring that one home. You will get 20 points for the transmitted data immediately, then 40 more when you get the second copy back to KSC. Total = 60. You actually get a little more total, because you got credit for two copies of the experiment, rather than just one.

I'm pretty sure it doesn't work that way.  (Am not in front of a KSP computer, so can't verify it directly at the moment.)

I believe the way it works is this:

• A given science result (i.e. "experiment A, done in situation B in biome C of planet D") has a certain maximum science value.
• That maximum is the total that you can get from it, regardless of transmission or physical recovery.

So if you have an experiment that has a cap of 50 points:  You can physically recover it, and get 50 points.  Or... you can transmit it for 20 points, then get another copy and physically bring it back, but you get 30.  Total value:  50, same as if you had just recovered it.   The transmitting didn't actually gain you anything.  (Didn't lose you anything, either.)  All that transmitting the science gained you, in that case, is that you got some of the science a little earlier, sort of like a payday loan.

Does that mean "never transmit"?  No.  I do it all the time.  Two reasons transmitting can be helpful:

1. for one-way missions
2. for getting some of the science earlier.  If I've just sent a manned mission to Duna, getting half the science immediately instead of waiting a year for them to come home can be useful in boosting me up the tech tree so I can launch my next, more challenging mission immediately.

One slight wrinkle to the above:  physically recovering a copy doesn't necessarily get you 100% of the cap value (depends on the experiment).  It may leave a little "dribble" left over, e.g. you only get 95% or something for a recovery, which leaves that 5% left over that you can obtain by recovering another copy later.  However, I generally don't bother mentioning that in discussions about science retrieval, since, 1. it's such a small amount that I find (for me, anyway) that it's not worth bothering with, and 2. it requires physical recovery-- transmission doesn't help you eat into that.

 

[bewing]

5 minutes ago, Snark said:

I'm pretty sure it doesn't work that way.

 

Ah, no -- because many experiments won't give you the maximum value with just bringing back one copy. You have to bring back three or four copies of an experiment to get all the last dregs of science points from a single experiment.

If you take a single copy of a SciJr reading on the launchpad, that's worth 7.5. If you go and take another copy, that's worth 1.6 more. The next is .4 and then .1. So the "Maximum" for two experiments is higher than the maximum for just one.

 

[Snark]

11 minutes ago, bewing said:

Ah, no -- because many experiments won't give you the maximum value with just bringing back one copy. You have to bring back three or four copies of an experiment to get all the last dregs of science points from a single experiment.

Yes, that's exactly what I was referring to when I said this:

28 minutes ago, Snark said:

One slight wrinkle to the above:  physically recovering a copy doesn't necessarily get you 100% of the cap value (depends on the experiment).  It may leave a little "dribble" left over, e.g. you only get 95% or something for a recovery, which leaves that 5% left over that you can obtain by recovering another copy later.

However, I'm pretty sure that transmitting doesn't let you tap into that "leftover dribble"-- the points you get for transmitting are taken out of the points you get from the first physically recovered copy.

So if you physically bring back N copies of an experiment (for any value of N, be it 1 or >1), you'll get a certain total amount of science for that, and transmitting doesn't help you.

In other words "Transmit, and then bring back N copies" gets you the same total science yield as "Just bring back N copies, without having transmitted."

TL;DR:  Transmitting never gets you extra science.  It just lets you get some of your science a little earlier, before you get home.

[bewing]

Well, then we shall have to test it, because I'm pretty sure it does.

Edit:

Just took a quickie to Gilly's highlands with a seismometer: points for returning one copy of the experiment = 360.

Points for transmitting one copy = 162, points for returning a copy after transmission = 212.7 = 374.7 total.

 

So the point is that a transmitted copy of an experiment counts as an extra half an experiment that you don't have to carry home with you.

Edited July 29, 2016 by bewing

[Magzimum]

Thanks for the feedback on my question. Much appreciated. I started putting a little less solar panels on my ships without running into problems, and I am quite happy that I started to look into this. 

As for the science transmissions, I will have a look into it. It is not too relevant for where I am currently in my career game.