Hey desht, you once told me you were interested in general feedback as to how the mod plays. So, since then, I have made the attempt to reproduce my experience as someone going into Pneumaticcraft completely blind, in text form.

It, umm, ended up getting somewhat long. So, story time! =P
https://docs.google.com/document/d/1Qo58V9WWgjYy4LHMLufUSdyTBbji3Zz-O5T8eu_9h2A

Note that this is really meant to capture the essence of what it felt like as I was playing it. At several points throughout this tale, you might be tempted to say "well you could just have done x". And yes, nowadays, I probably would do x. I probably discovered x myself by now. But the point is describing how discoverable the mechanics are for someone with zero prior exposure to the mod. And, of course, this is only one single data point. A different player, with a different playstyle, might learn things in a different way.

But I hope that it helps you anyway.

As a TL;DR, based on my experience, I would like to make the following suggestions to smooth out the earlygame:

• The gas lift really needs to be linked to from relevant pages on the ingame wiki, for example the entry on oil.
• Thermal mass of blocks could be looked at. The paint mixer very clearly has completely the wrong value. The other conductive blocks, meanwhile, have almost no thermal mass worth mentioning; they could use a little more. Lava, too, should last longer than a few seconds before turning into obsidian.
• The refinery would probably work much better if it was reworked. For example, you could have one "refinery controller", which is heat conductive, offers the interface, and acts as the oil tank. It would also be the expensive diamond-gated craft. Then, you'd have up to four "refinery column" blocks which you stack on top of the controller. These would be a lot cheaper than the controller, would not be heat conductive, and each would be the output tank for one particular product (much like the refinery works now). Such a configuration would be much easier to interact with heat-wise, which is kind of important because the refinery is the first heat-conductive machine that a new player encounters. It needs to allow the player to easily discover the mechanics, instead of being one of the most difficult to deal with heat machines in the mod.
• There is a bit of a cognitive disconnect between the heat sources available to the player and the temperatures the player expects to achieve. I ended up being so frustrated at one point, because with all the fire and the lava in the world, it seemed impossible to hit a mere 100°C. That's just boiling water! How can lava fail at boiling water? If, on the other hand, the target temperature had been 1000°C, it would have made sense to be difficult to achieve. I mean, I would still have been somewhat frustrated, but I would have understood it. Obviously, you may have had perfectly good reasons to choose the temperatures you chose, and I don't know those reasons; I can only report how they made me feel a bit incompetent.
• You can probably remove torches as a heat source altogether; they literally have no positive or negative effect whatsoever in the current heat mechanics that I could find (and I tried!). So you might as well get rid of the extra processing overhead, as well as creating the false expectation that the player is to be able to utilize them in some way, or needs to be careful with where to put them.

That's some great information right there, thanks. I will take a little time to process it all and post some answers here. It does sound like a lot of your problems could be alleviated by some better documentation and guidance within the mod, which is something I definitely want to work towards. E.g. 0.8.1+ has a lot more advancements to help guide the player.

Not sure if you've read #130 which covers a lot of rebalancing and reworking ideas, many of which are quite fundamental and thus likely to wait until the 1.13 version ("PneumaticCraft 2"?). You'll probably find it quite interesting.

A few initial comments:

• Agreed on docs for the Gas Lift. I'm adding a paragraph to the Oil IGW page to specifically mention using the Gas Lift and Liquid Hopper for pumping and moving Oil around. As for its name, it is based on a real-world device which uses pressurized gas to pump liquids out of the ground, but maybe it could be called a "Gas Lift Pump" in Minecraft for ease of JEI location...
• The Plastic Mixer does indeed have a very high thermal capacity, but note that you don't need to heat it at all if you're solidifying liquid Plastic. It only needs heat if you want to melt down solid plastic back to the liquid form (which is uncommon but possibly useful if you have too much of a colour you don't need). That's mentioned in the IGW docs (first comment in fact :-)
• Overall thermal properties of blocks might benefit from a review, but that's a significant balance issue which means probably not until 1.13. In general, machines should have a low thermal resistance, though so they don't take to long to heat up.
• You're right that torches don't do much: they have a high temperature, but extremely high thermal resistance, so their heat moves very slowly.
• All thermal blocks lose some heat to the air but the Refinery is most noticeable given the number of faces of the block. However - and this probably needs better documenting somehow - that heat loss is completely negated by placing any non-heat-aware block beside each face of the refinery. Doesn't have to be a full block, either - trapdoors work as covers, Chisels & Bits decor would work...
• Regarding a Refinery rework - possibly, but definitely not before 1.13. Your idea would be quite intuitive though, I agree.
• Regarding how fast Lava gets used up as a heat source: a few seconds is too fast, yes. I'm experimenting with some values here, but try setting D:fluidThermalResistance=500.0 in pneumaticcraft.cfg. The default is 10.0, which does feel much too low. And due to a bug, flowing lava does in fact get added with a resistance of 500.0, which explains why it worked so much better for you. Static Lava should be the same, really. 500.0 is arguably a bit high, since a bucket of lava lasts ages then.
• (By the way, if you have Thermal Foundation installed, try using a bucket of Blazing Pyrotheum...)
• Liquid Tanks: the Liquid Hopper can also be used as a general purpose tank (fluid pulling and pushing can be disabled with the redstone settings). It's very cheap to craft, keeps its contents when broken and holds 16000mB of any fluid. I added a note in the IGW page about Oil about using that in the early game.
• Air Grate IGW does indeed need updating to remove plastic plant references; the information on Heat Sink cooling is accurate, though. I also don't like the way the grate is omnidirectional instead of only affecting entities in the direction it points, but changing that behaviour might wait till 1.13.
• The Plastic Mixer does need all of red, green & blue dyes to make any colour of plastic (except white). This is a bit odd, yeah, but hey :) Technically (since we're mixing paint, not light) it should need cyan, magenta, yellow & black dye, but that's probably a bit OTT. Still, maybe for 1.13. Then again, we may well get rid of the concept of coloured plastic entirely for 1.13 anyway since it doesn't add much to gameplay (it was more interesting when plastic plants were a thing since you had to farm all the plants). That issue I linked up above has some more discussion about that.

Alright, I had some time to play with 0.8.2 and the changed fluidThermalResistance setting. At first it confused me, because I thought you had increased the actual thermal mass of lava - the amount of heat it can give off before turning to obsidian. But if I understood right, the thing that actually changed was the heat transfer out of lava into other conductive blocks, right? Since the lava now doesn't transfer everything at once, in a crazy spike to multiple hundred degrees, it takes longer for it to turn into obsidian, and a more gradual heat supply is easier to make use of.

So far, so good. The thing is, I always ask the same question: "but can it run Crysis the refinery?" This is a key question, IMHO, because that's the moment where lava is the relevant heat source. After you have jumpstarted the refinery, you have some gasoline and kerosene to work with, thus making liquid compressor driven vacuum tubes viable. In my personal playstyle, I see very little use cases for running machines off of lava besides the refinery jumpstart.

And with fluidThermalResistance=500, I'm afraid the answer to that question is "no". It cannot run the refinery. Upon surrounding the bottom refinery block with lava on all four sides, the temperature will reach 99°C, and then more or less stay there forever. If you are lucky, you might get one 10mB processing step per minute. And sure, the lava lasts a very long time like this, but what you have built here is essentially just a giant heat sink. Given enough time, the large refinery with its many open faces simply dissipates all the heat energy contained in four blocks of lava without doing any work whatsoever. Obviously, if I start submerging the entire 4-block tall structure in lava top to bottom, or enclose the upper layers in cobblestone, then yes, it will run. But that is awkward and impractical, and it will not be something a new player thinks of trying.

So this is basically the other extreme. The previous setting was not useful because all heat was dissipated to the environment instead of doing useful work because it was transferred too quickly. The current setting meanwhile is not useful because all heat is dissipated to the environment instead of doing useful work because it is transferred too slowly.

So I experimented with other values. In the end, I think, 200 to 250 is decent. Upon using two buckets of lava while the other two bottom side faces are not open (apparently those are more important than the open faces further up), the refinery will actually got to 120°C and do real, useful work for a useful amount of time, processing more than half a bucket of oil until the lava runs out of heat.

Now, in regards to the plastic mixer...

Whale Isle Beef Hooked. Between JEI showing all plastic recipes as requiring heat, being unable to get any plastic out due to missing dye during my first experiments with the block, and the big red "this machine requires heat" info box in the GUI, it neither occurred to me nor did I accidentally chance upon the concept that there might be a difference between melting plastic items and using already liquid plastic from the internal tank! Wow, I feel so stupid now, having spent a full ten minutes on heating up the darn thing every single time that I needed plastic so far. But at the same time, I feel incredibly glad that I no longer need to to that. This will decrease my kerosene use by approximately 95%. Conservative estimate.

I'll give #130 a read later, it looks like there's a lot there to digest.

I don't think it's impractical to clad every face of the Refinery. As you've come to realise, it's pretty much a necessity. Just be aware that you can use any non-conductive and non-air block. For example, trapdoors actually work pretty well (and they can be opened - without losing heat - if you need access to the refinery blocks behind):

That setup will process (from cold) 2360mB of oil with one bucket of lava, which is absolutely loads. More than enough to get some high quality fuel to run a couple of liquid compressors and a vortex tube.

So I don't think there's any problem here. The thermal resistance of fluids is configurable, so you're welcome to change it, but 500 is a good default IMHO.

And of course, much prettier setups than this exist, e.g. this was posted on Reddit just today: https://www.reddit.com/r/feedthebeast/comments/9u4dlz/build_showcase_anthony_kanes_oil_refinery_mark1/ - including a fully cladded 4-block Refinery. It's from 1.7.10, but the heat mechanics haven't changed.