LCOE is one of the most commonly used (and abused) metrics within energy-related articles, websites, presentations and such like.
The US Energy Information Administration describe it as representing “the per-kilowatthour cost (in real dollars) of building and operating a generating plant over an assumed financial life and duty cycle”. You’ll see a variety of differently worded definitions however, along with a variety of simple to very complicated-looking calculation methods too.
The basic concept is pretty straightforward: add up all the costs accrued in building and operating a power plant during its lifetime ($) and divide by all the energy it produces during that lifetime (kWh or MWh); thus producing a $/kWh (cost of energy) number. When adding them up, a discount rate is used to reduce the present values of both costs and energy which are accrued in the future.
By looking at lifetime costs of energy, LCOE is useful in comparing two ways of generating energy which may have very different cost structures. For example solar may look expensive today compared to gas, in terms of the cost to build an amount of capacity. But it has few operating costs (sunshine is free), whereas a gas plant has a lifetime of spending money on gas to run it. So which is cheapest overall? The answer obviously depends on a bunch of assumptions: how sunny is it? How often does the gas plant operate? What happens to the price of gas in future (and to what extent are you discounting those future costs)?
The numbers also depend on the details of the calculation. Some simple calculations don’t incorporate factors such as taxes or depreciation or interest rates, whereas some do, and choices of discount rate will differ. The results of different approaches to the same question can be wildly different. While it can be informative to compare LCOE results (“energy costs”) for different power options when they’ve been calculated in a single study using a consistent methodology and assumptions, it’s a fool’s game to try to compare numbers derived by different authors for different purposes in different studies, without clear knowledge of how they were derived in each case.
To quote the EIA again: “while LCOE is a convenient summary measure of the overall competiveness of different generating technologies, actual plant investment decisions are affected by the specific technological and regional characteristics of a project, which involve numerous other factors.”
It’s also important not to confuse levelized costs with energy prices (i.e. costs to energy purchasers and consumers). LCOE is an attempt to quantify the underlying cost of generating a unit of electricity, but doesn’t account for other costs outside the power plant gates (transmission and distribution, taxes, environmental charges, distribution and retail profits and so on). It also says nothing about the value of a unit of electricity; which determines its price and which depends on a variety of other market factors, for example when it is generated (relative to demand).