LCOE — Levelized Cost of Energy

Typical LCOE Graphic

Levelised Cost of Energy is frequently presented showing renewables are cheaper. And it’s true, so far as it goes.

The definition of LCOE is the total of all costs over the life of the asset, divided by the amount of electricity produced over the same life. Sounds good? But wait, what discount rate are you using? What availability and utilization do you assume? These assumptions are not transparent. Different analysts use different assumptions.

But there’s more costs to electricity than the cost at the terminals of the generator, which is what LCOE measures.

The power bill you get at home is ~35% generation cost (“LCOE”), 50% transmission cost, and 15% overheads and profit.

Storage

Levelised Cost of Storage (LCOS). In my state, Queensland, to buy a Tesla powerwall 2.0 is about $17,500. Tesla guarantee it for 2800 cycles, and it stores about 13 kWh. So that’s a lifetime storage of 36,400 kWh. You can do the maths in your head, that’s about $0.50 per kWh LCOS. (Ignoring finance costs, which may double it)

Bear in mind:

1. Your cost of your own solar power is around 5 cents / kWh
2. You can buy it from the grid for 25 cents.

Hence, the Tesla powerwalls haven’t exactly taken off.

A Powerwall at home will return around 92% of the energy you put in. On the other hand there are (very early) at-home hydrogen based solutions available that return only 50%, but end up being cheaper overall, due to how much bigger you can make them and how much longer they last.

It’s likely to be economic to build a 15kW solar and H2 storage system, run two hydrogen cars and never have a power nor fuel bill again. That may end up more economic than batteries, since fuel cells not only last at least twice the life of a battery, but they cost around $100/kW, so a $5000 fuel cell is cheaper than a $18,000 battery.

But aren’t batteries going to get cheaper? What of Elon’s promise that batteries will cost only $100 / kWh by the middle of the decade? Well, over 2022 they went from around $400 in January to $480 in December — based on Tesla’s Megapack pricing. (USD). And that’s just the battery cost. Allow more for the site civil and electrical connection works, usually around another 50–80%. Yes, the spike in battery prices this year may be due to short term factors but it’s a brave predictor who will say they will drop, or that some other ‘short term factors’ won’t keep occurring.

Pumped Hydro storage is cheaper than batteries. The cost overruns on Snowy 2.0 are disappointing but this asset will have >100 years life with good maintenance and occasional mech refit every ~30 years. Just bear in mind you lose 20–30% of the power in the pumping operation, and another 10% in the transmission to and from.

Local community batteries help a bit but not much since the problem is local weather. Grid scale batteries are better, but suffer transmission losses which can reach 20%.

Transmission & Distribution

T&D costs are also excluded from LCOE. (Remember that’s 50% of your power bill at home)

In a renewables-heavy grid, transmission cost goes up because power is shifted all around Eastern Australia from where it’s sunny or windy to where it isn’t. Hence the ‘rewiring the nation’ project for ~$100B by the time it’s finished.

So even if in a renewables-heavy grid, generation cost falls to zero, transmission will increase, and may even double.

Curtailment (wastage)

In Australia’s “NEM ISP” — National Electricity Market Integrated Servcies Plan, the end state in 2050 assumes that 20% of the electricity generated is ‘curtailed’ or ‘spilled’, that is, wasted. The reason for this is that it’s easier to overbuild the variable renewables than to build extra storage. Again, this 20% is excluded from the LCOE calculations you have seen.

Backup

The NEM ISP plans 10GW gas-fired peaking plants (around 20% of the dispatchable ‘turn on right now’ power). This is an asset that will be sitting there, used only occasionally. It will incur all the finance and most of the maintenance costs, but will be used only rarely. Which is lucky, because it’s peaking open-cycle, which inefficiently wastes all the heat.

LCOE Summary— what they don’t include:

So, in summary, LCOE as commonly presented, is a deeply flawed method of measurement. Here’s what they don’t include:

· Storage

· Transmission

· Curtailment

· Backup

All of which adds up to a lot more than LCOE by itself. And it’s very hard to calculate any of these in the same way.

They don’t tell you the financial assumptions such as discount rate (cost of money, interest rate), cost of CO2 offsets etc.

So what’s better?

Surely we can come up with a good measure of electricity costing? Yep, we certainly can, the good old “NPV” or Net Present Value. This can cover the entire cost of the grid, with all operations, maintenance and fuel costs. The finance assumption is transparent, because you always know what rate is used, commonly around 7%.

When you apply NPV models to the entire grid designs with various technologies, the clear winner in Australia is Combined Cycle Gas. New designs are available with emissions only 2.5x that of renewables, only a quarter of coal.

Even when you pay $120/t to offset the extra CO2 emissions, a network heavy in Combined Cycle Gas with say 30% renewables is a cheaper solution for Australia. It will cost less than half the current proposal in NPV at 7%. (Remember they won’t even tell you what % LCOE uses unless you ask them.)

Europe has also realised this. On 6th July 2022, the EU parliament published a new rule that gas plants with emissions at 270g/kWh would be classes as “Green”. (Compare 870–920 for black coal, 1200 for brown, and 100 for renewables).

This tech can be rolled out now. The best solutions in Combined Cycle Gas achieve 285g CO2/kWh out of the box, and can be improved with partial hydrogen and better use of waste heat, easily beating EU criteria.

Yes, gas prices are costly in Australia, but that is due to nonsensical policies and the fact we export 75% of our production as wasteful LNG. (Liquifying it consumes 40% of the energy + more for tansport). Sensible policies would see gas prices fall to $10/GJ, which translates to around $55/MWh in fuel costs for a state-of-the-art combined cycle gas powerplant meeting EU criteria.

Summary: LCOE is one of six cost elements in your power bill

1. LCOE (Generation cost, currently around 35%, will fall with renewables)
2. Transmission and Distribution (currently around 50%, will approximately double with renewables, maybe more)
3. Profit and overhead (currently around 15%)
4. Storage (currently zero, hard to predict in future. We’re paying $10B+ for Snowy 2.0 which is less than a fifth of what we need)
5. Curtailment / wastage (currently zero, predicted to be 20% of LCOE by 2050)
6. Backup generation, eg gas (currently included in LCOE, in future we plan to keep 10GW available, which will be an additional capital and maintenance cost even if not used.)

Your energy costs are therefore going up, whatever any politician says. Unless you can use behind-the-meter power which eliminates much of these costs. As usual, the rich with their own houses and solar panels will be at least partly insulated from these increased network costs.

So next time you hear a talking head saying solar costs 3–5 cents per kWh, try testing the speaker’s knowledge that such figures are “LCOE” and ask them if they understand that excludes the cost of storage, transmission, backup and and wastage, and that the correct way to work out the cost of electricity is to apply NPV techniques to the entire grid design.

Our politicians need to be honest and tell us the real cost of the renewables-heavy path.

PS: 2023–03–08 What a surprise, Copper String (powerline from Mt Isa to Townsville) now costs $5B not $2.5 last year or $1.5 as it was before the pandemic. What do you think the final cost will be? This is an excellent project with the ability to attract lots of business and growth along the 800km Flinders Highway but what confidence can any of us have in the cost of this, after Snowy 2.0 and the Gabba?

Interesting references:

https://www.eia.gov/outlooks/aeo/pdf/electricity_generation.pdf

Levelized cost of electricity - Wikipedia

The author is an engineer with a post-grad diploma in renewable energy (2018) and has been following this stuff since the 1990s, back when it was called “Alternative Energy”.