Power Plants come in lots of different flavors. The early ones are just terrible, yes, but you’ve got to buy them to get off the ground. Almost all of the other ones are worth it at some point or another, but sometimes certain plants are better than other ones. How do you tell the difference?
I’d like to categorize power plants in two ways; by judging their efficiency and their throughput. Some power plants fall into some other categories (1 and 2) but the rest of them fit on a sliding scale of efficiency vs. throughput. Not all plants are created equal, but knowing where on that spectrum you want to be at a given point in the game can be valuable.
1. Really low throughput: Any plant that powers three cities or less falls under this category. These plants just don’t power enough cities to be more than a tactical play at the begininng of the game. If you’re considering buying one of these after Turn 3, consider reconsidering. If it’s actually a good idea to do this, then you’re in a pretty bad situation to begin with. It’s a good idea to think of the 21 Plant (2 Hybrid -> 4 Cities) and the 24 Plant (2 Garbage -> 4 Cities), or even the 28 Plant (1 Nuke -> 4 Cities) in this group as well, since they are seldom efficient enough to justify their low throughput.
2. High Efficiency/High throughput: They way I’m defining these things, only two plants consistently fit into this category: the 50 Plant (nothing -> 6 Cities) and the 39 Plant (1 Nuke -> 6 Cities). You won’t see the 50 Plant until Step 3, and you’ll almost never see the 39 Plant until then either. You’ll be powering six or more cities while paying next to nothing for resources, so these are always strong plants to have (unless you absolutely need something that powers 7 cities). Unfortunately, you won’t get to use them more than one or two turns at most. Oh well. Sometimes the 30 Plant (3 Garbage -> 6 Cities) or the 38 Plant (3 Garbage -> 7 Cities) fit this role as well, but only if Garbage is super-cheap at the end of the game.
3. High Efficiency/Low Throughput: Plants that only power four Cities (sometimes five) but are very efficient. The prime example is the 29 Plant (1 Hybrid -> 4 Cities); this is one end of the interesting spectrum.
4. Low Efficiency/High Throughput: Plants that take three resources to use, but will power six or more cities. This is the other end of the interesting spectrum.
So which is better, efficiency or throughput? The answer: throughput. What, you thought it was going to be more complicated than that? Of course it is! But powered cities are victory points, and you want to make sure you get as many of those as possible, since that’s how you win. Given a choice, I’ll take throughput over efficiency most of the time, but that’s not all of the time, and sometimes I don’t get the choice.
If you’re going to go for efficiency, you want that to pay for itself. So, the more turns you can use that plant, the better off you will be. This means that efficiency is really only valuable in the early game, and gets more valuable on a map with more expensive resources and connections. Efficient plants also set you up nicely for a Step 2 Stall.
Be careful, though, because having a low-throughput plant takes away options in the endgame. In the beginning, I try to aim to power one more than the number of cities needed to end the game (4 players -> 18 cities, 5 players -> 16 cities, 6 players -> 15 cities), so a plant that only powers 4 cities either needs to be replaced at some point, or you have to get some higher-throughput plants later in the game to compensate. A lot of the time this can outweigh the benefit of the efficiency you get early game in terms of dollars, so try to make your early dollars count more than the late dollars you may lose.
The other reason you’d want to value efficiency is if you fear that resources may run out. It’s very hard to get choked out of using an efficient plant. However, using only efficient plants will rarely get you enough cities to win the game, so you’ll have to spend some resources at some point. Smart positioning within resource markets can also alleviate this issue, sometimes better than efficiency can: if I’m the only person burning garbage, it doesn’t matter how much I burn (barring some really weird stuff) because nobody else can choke me.
Needless to say, on fast/cheap maps, there is no reason to be efficient. Power plants can just be evaluated by the number of cities they power.
Summary: High-throughput plants are long-term investments, and they are risky. In a game with at least 4 people, this is what is required to win. Efficient plants can be used as a tactical play on expensive maps to get a quick lead and try to hold onto it, but otherwise I view it as a middle-of-the-pack strategy.