Good questions, Armchair Critic.

I didn't know of the exact answer, so I looked it up to avoid looking like Try McClure in this scene of The Simpsons:

http://www.youtube.com/watch?v=wCOirUVBFW0&t=61s

Yes, labour costs matter. Lower-labour-cost countries can build infrastructure more cheaply, and that applies to all energy tech, not just renewables. It's one reason China can roll projects out so fast.

But labour isn't the whole story. Wind and solar are mostly capital-heavy rather than labour-heavy, so labour costs tend to change how cheap projects are, not which ones come out cheapest overall.

The "why not let China build it all?" angle is really an industrial policy question. We already accept that trade-off with manufactured goods. Energy's different because governments sometimes choose to pay more for local capability and control, and they make that choice whether the generation is renewable, fossil, or nuclear.

And the offshore solar project is interesting mainly because it shows how much innovation and scaling is still happening, not because it breaks the basic economics.

The costings can all be found across resources from the IEA, Lazard, and CSIRO. LCOE is a good starting point.

Will this nonsense ever stop ? There are no renewables at this stage !

""Cheap to build" is a claim about marginal costs,"

Oh so now we see JD agreeing that renewables are only cheap if you look at a small component of the picture. Cheap to build... yes. Cheap to incorporate into the grid...not so much.

"The economic fact remains that diesel used in fossil fuel production is effectively untaxed relative to other energy inputs. "

Well the diesel used to construct the various renewable facilities is also untaxed... but we won't call that a subsidy because ... reasons.

"unless retail prices fall immediately."

When the current government came into power they promised that retail prices would fall because of their policies in increase renewable penetration. They are still making that promise although the timelines have necessarily been extended.

So retail prices were the measure until they proved the claims were false and then suddenly they no longer counted. And many continue to fall for it.

No, that's not what I'm saying, mhaze.

//Oh so now we see JD agreeing that renewables are only cheap if you look at a small component of the picture.//

I'm simply restating the original claim precisely.

"Cheap to build" is a marginal cost claim. Calling that a “small component” doesn’t refute it, it just rejects the legitimacy of partial analysis unless it answers every downstream question.

//Cheap to build… yes. Cheap to incorporate into the grid… not so much.//

Those are different questions answered by different analyses.

Treating grid integration costs as a veto on marginal build costs collapses distinct metrics into one after the fact. That move has already been addressed.

//Well the diesel used to construct the various renewable facilities is also untaxed…//

Diesel used briefly during construction is not economically equivalent to diesel used continuously as an input into fossil fuel extraction and processing. One is incidental and finite, the other is structural and ongoing. So collapsing them into "both untaxed" erases the distinction that actually matters.

//When the current government came into power they promised that retail prices would fall…//

That was a political forecast, not an economic definition. It has no bearing on marginal build costs, and delayed outcomes don’t retroactively falsify LCOE or auction data.

//Retail prices were the measure until they proved the claims were false.//

Retail prices rose due to gas shocks and network costs. Evidence suggests they would have risen further without renewables already in the system. Either way, retail outcomes don’t negate marginal build-cost data.

At this point, the disagreement isn't about evidence. It's about whether different questions can have different metrics, or whether retail prices are being treated as a universal veto once outcomes disappoint.

JD what you’re doing here is repeatedly asserting that a partial metric remains valid simply because it is internally consistent, while ignoring how it is being used in practice.

No one is denying that “cheap to build” can be defined narrowly as a marginal cost claim. The objection is that this narrow claim is routinely deployed to imply broader conclusions about affordability, transition efficiency, and consumer benefit. Once that move is made, it is no longer enough to say “this answers a different question” and treat downstream costs as analytically optional.

Grid integration costs are not a retrospective veto; they are a foreseeable consequence of scale. When a technology’s marginal attractiveness increases precisely because system costs are externalised, those costs cease to be merely “downstream”. They become part of the economic reality of deploying that technology at meaningful penetration.

On diesel use, the distinction you draw is one of degree, not kind. The point is not that construction diesel equals fossil fuel combustion, but that selective exemption is being invoked to dismiss one case while normalising another. That is still an accounting choice, not a neutral clarification.

Dismissing retail prices as politically irrelevant also misses the point. They are not being used to falsify LCOE line by line, but to test whether narrowing the metric continues to track outcomes that were publicly promised on the back of those same cost claims.

The core issue isn’t whether different questions *can* have different metrics. It’s whether insisting on ever-narrower metrics once results disappoint is an act of clarification, or an attempt to protect a claim from the consequences it was always meant to support.

Graham,

You're now arguing about how a claim is sometimes used, not whether the claim itself is true.

I've been explicit throughout about scope. "Cheap to build" is a marginal generation-cost claim. If others overextend it rhetorically, that's a misuse, not a reason to declare the metric illegitimate or to treat all downstream costs as analytically prior.

System integration costs are foreseeable and modelled, which is why they appear in system plans and not in LCOE tables. A cost being foreseeable doesn't collapse all analytical layers into one, nor does it make marginal cost meaningless.

As for diesel, duration and recurrence matter. Incidental construction inputs are not economically equivalent to ongoing production inputs, and calling that distinction "accounting choice" doesn't make it arbitrary.

And retail prices test policy outcomes, not the validity of generation-cost metrics. Political promises don't determine economic definitions, and disappointed outcomes don't retroactively change cost rankings.

It seems the disagreement now is about whether a bounded claim can remain valid once others try to load it with conclusions it was never meant to carry. I think it can. You appear to think it can't. That's the crux.