Colony Efficiency

[Catbriar Saissore]

After reading and rereading numerous sources of PI information, I can set up PI on a planet and get all the structures
connected. I follow the connect the structures logic. It was a struggle and the information out there is overwhelming and in many cases, I think outdated. That said:

So, pretend I have a colony set up as follows:
Extract Base Metal --> Storage #1 --> Process Base Metal into Reactive Metal --> Storage #3
Extract Micro Organisms --> Storage #2 --> Process Micro Organisms into Bacteria --> Storage #3
Storage #3 --> Process Reactive Metal and Bacteria into Nanites --> Launch Pad

A very basic set up. So I start extracting (say 4 day cycle) and go away for two days. When I come back, Storage #1 is
overflowing because the extractors produce more metal than processing can handle. Storage #3 is full and throwing out
Bacteria because I am producing more Bacteria than Metal. Basically a very inefficient operation. As I say, I am making
this up but you get the idea.

Somewhere in this incomprehensible PI user interface there has to be information to initially set up an efficient colony
where you are not losing product or creating bottle necks. I want to be proactive up front not reacting later. (Is there a
simulation mode?) I do not expect anyone to explain this in a post, that would be cruel and unusual punishment. Would you
kindly point me to a good video or article that gets into the detail on efficiently setting up a colony. Thank you.

[Troven Smalvard]

as i remember basic factories consume 6000 units of p1 /hr, so depending on how many factories you have you can scale your extarction to meet this average, it displays it in the window when your 'programing' and can move the heads around.

that said, you'll see the graph in same set up window that it harvests on a decreasing curve so although the average is x the peak at the start will be much higher, which is why you buffer as it looks like you are doung with your storage 1.

the trick is to gauge how much buffer you have at the start of the cycle. i'm sure you could math it but can just use an app like neocom to show if you'll have losses

i'm not sure if this helped? feel free to hit me up in game, think would be easier to talk it through on mumble with your stuff on screen

[Catbriar Saissore]

Thank you Troven. Yes, 6000 per hour makes sense. The Uni wiki says to get average extraction to 12,000 per hour. Then it shows two factories for the extractor, ties in perfectly! The numbers are out there, it's just putting them all together.

[Atomsk Agittain]

PI gets odd, because your sliding scale of p0/h changes based on how long you set your pulls for. At one day I believe the extractor cycles every 15 minutes whereas the factories cycle on the half hour. On the other hand a max length extraction cycles every four hours and pulls massive amounts up in the beginning so you can easily overload your system at the start and lose production at the end. Or it did before the PI update, I haven't messed with it since then.

Keep in mind that the distribution of p0 changes periodically as well so what works today might not function tomorrow. (I had 10 factories on one planet but the extractor couldn't always meet demand)

[Catbriar Saissore]

Thank you Atomsk, your point on extractors pulling in more P0 at the start of the cycle is well taken.

Let me share what I have found:

I am Level IV on Planetary Interaction. With the following structures you pretty much use up most of the CPU from your Command Center:

One Command Center
One Launch Pad
Two Advanced Processors (P1 -> P2)
Four Basic Processors (P0 -> P1)
Two Extractors (4 heads each)

The following math is in units per hour not in units per cycle. It is much easier to think in units per hour because your extractors display units per hour when you set them up.

A single advanced processor requires 80 units of input per hour to operate continuously (Two requires 160 per hour)
A single basic processor requires 6,000 units of input per hour to operate continuously (Four requires 24,000 units of input per hour and outputs 160 units per hour)
A single extractor with four heads on a 4 day cycle can generate 12,000 units per hour on average. (Two extractors generate 24,000 units per hour on average)

You can follow the math and see how the numbers tie together. This also ties to the 12,000 units per hour mentioned in a prior post.

What I did not address is storage. I tried to use the launch pad as the only storage which was a mistake. Because the extractors pull far more than 12,000 units per hour when first started you will need at least one storage unit to act as a buffer.

I hope that helps some other Unista starting out in PI.

[Christoph Patrouette]

In my experience it is much more efficient to only have 1 extractor per planet and to then alternate between the input materials. It takes a bit of storage, but that is easily given by the saved CPU and PG. It also allows more flexibilty to reach far better material positions on larger planets.

[Mike Kingswell]

What I did not address is storage. I tried to use the launch pad as the only storage which was a mistake. Because the extractors pull far more than 12,000 units per hour when first started you will need at least one storage unit to act as a buffer.

Just checked up on that and am a bit confused now (shortly after a nap for me ^^').

A p0 needs 0.01 m3.
a Launchpad has 10000m3

It would take a LOT of extracting over 6000/h (12k/h if you use 2 factories) to fill that up.
In those rare cases I just shift some p0 into the poco.

Am I doing something wrong/missing something?

In my experience it is much more efficient to only have 1 extractor per planet and to then alternate between the input materials. It takes a bit of storage, but that is easily given by the saved CPU and PG. It also allows more flexibilty to reach far better material positions on larger planets.

On how long a cycle do you keep it then? You don't want to run out of p0(1) while extracting some p0(2) and have your factories not running?

[Atomsk Agittain]

In the beginning the cycles are long enough and the pulls large enough that the factories can't handle the initial loads so they back up and by day 3 or so of a week pull there is more p0 than the launchpad can handle.

Let's say I have one extractor and five basic factories. I set up for a week run and have enough extractor heads that overall I have 30k/h needed to run them.

Pull starts, factories idle for 4 hours while the heads run.
First pass extractor pulls up 200k, 2/10km3 space is used, factories then pull out 30k/h for the next 4 hours while the extractor cycles.
Extractor dumps 180k into the launchpad 80k remains from last round so I now have 2.4k+(8x5 rounds of p1)/10km3 in the launch pad.

So unless you math it out or check in on it, which on a week pull I don't want to mess with it, that is how it overloads.

[Mike Kingswell]

Thanks for clearing that up.
Didn't know what you were saying since I run on one 3day one 4day cycles per week.

[Xaviar Onassis]

I find the Neocom II app on my iPhone (not sure if there's also an Android version) incredibly useful for PI tracking. It shows you (among other things) a plot of your storage capacity over time and where you're going to be losing extracted products due to lack of storage.

[Catbriar Saissore]

Thanks Atomsk, you answered Mike's question better than I could. I was going by instinct rather than math.