The Hidden Costs of "Cheap"

Here is the concession the other side never expects: solar and wind really are the cheapest power per megawatt-hour. That number — LCOE — is also the most misleading figure in energy policy, because it prices a single generator in isolation and quietly drops everything it takes to turn it into a grid that works at 3am in still weather. The drops, layer by layer:

The evening gap

The reliability problem in one chart. Drag the slider: more solar deepens the midday glut, but the demand peak lands at 6–7pm — exactly when solar is gone. That gap doesn't shrink with more panels; it has to be filled by something firm. Layer 1 of the hidden bill below.

Solar in the mix70%

solar generation total demand net demand — what gas, coal, hydro or nuclear must still supply

At 1pm, solar is pouring out ~19 GW — push the slider and watch the midday belly sink toward zero (that surplus gets curtailed). But at 6:30pm, when demand peaks, solar is already ≈0: the evening gap stays at ~31.2 GW no matter how much solar you add.That residual peak is the firm capacity wind and solar can't deliver — and the cost LCOE never shows.

Representative high-solar day, NEM-scale (illustrative — the curve's shape is well documented by AEMO & CAISO; exact MW vary by day and season).

Firming

+A$6–8/MWh

LCOE prices a panel as if the sun always shines. It doesn't, so the system needs storage, gas or hydro to fill the gaps — AEMO's ~74 GW / 646 GWh dispatchable fleet by 2050. None of that is in the generator's LCOE.

Transmission

+A$10–18/MWh

Remote wind and solar zones need thousands of kilometres of new poles, wires and interconnectors (Rewiring the Nation, HumeLink, Marinus). Those projects have 2–4× their first estimates. A coal plant sat next to its load; a renewable grid is spread across a continent.

Overbuild & curtailment

hidden in capacity factors

To deliver firm energy through low-wind, low-sun weeks, you build far more nameplate than you can use — and spill (curtail) the surplus on good days. The wasted megawatt-hours still had to be paid for, but they don't appear in a sunny-day LCOE.

Replacement

every ~25–30 yr

LCOE amortises over a panel's life — but a renewable grid is rebuilt two to three times over the 60-year life of a single reactor, each rebuild a fresh round of capital, material and waste. Long-lived firm plant spreads its cost over far more energy.

System security

+A$2–6/MWh

Spinning coal and gas turbines gave the grid inertia and fault current for free. A converter-based grid has to buy those services back — synchronous condensers, system-strength remediation, frequency markets — another line LCOE never sees.

Generation cost ≠ system cost

Add the layers and CSIRO GenCost's own integration cost reaches ~A$16–28/MWh by 2030on top of the headline LCOE — before transmission. The honest claim isn't "renewables are expensive." It's that a cheap generator can sit on an expensive system, and the bill the public actually pays is the system, not the generator. That's the number this whole site exists to count.

See it move in the simulator →The firming bill in full →