The cheap-solar story is running out of road. IRENA’s 2026 cost report says global utility-scale solar sat at about $44 per MWh in 2025, onshore wind at roughly $33, and offshore wind at around $78. Those are still respectable numbers, but they no longer rescue a sloppy project.
For years, developers could hide a lot behind falling hardware prices. A site with awkward access, a weak grid connection, or a mediocre contract could still look fine if panels were a little cheaper this quarter. That trick is fading. The next winners in renewables will be the firms that know how to finance, place, operate, and contract a project as one system, because the module is no longer doing all the heavy lifting.
The price curve has flattened
The most useful thing IRENA has done in its 2026 report is puncture the fantasy that solar and wind will keep getting dramatically cheaper on autopilot. The direction of travel is still down, but the slope has softened. Solar utility-scale projects averaged about $44 per MWh in 2025. Onshore wind sat near $33. Offshore wind, which has more moving parts, more steel, and more headaches, averaged about $78.
Those figures matter because they show where the easy gains have already gone. In the 2010s, panel costs fell so hard that developers could sometimes make almost any half-decent project work. Manufacturing scaled, supply chains matured, and the industry harvested a long run of efficiency gains. That phase is ending. Polysilicon is not doing the old heroic price-collapse routine, steel is not falling in love with developers, and labour is not getting cheaper because a power point deck says it should.
I think this changes the whole tone of competition. The game is no longer, “Who can wait longest for the next cheaper module?” It is, “Who can build the best overall machine around a fairly fixed piece of hardware?”
Finance is now part of the technology
Once equipment prices stop dropping fast, the cost of money becomes far more visible. A project with cheap debt, long repayment periods, and a credible buyer for the power can beat a technically similar project that looks fine on paper but scares lenders.
Long-term power purchase agreements matter so much for this reason. A 10-year or 20-year contract with a creditworthy corporate buyer turns an ugly revenue forecast into something a bank can underwrite. Amazon understood this early and kept piling on deals, announcing 13 new renewable energy projects globally in 2023 and pushing its portfolio past 500 projects. A giant buyer can smooth cash flow, and smooth cash flow is what lenders actually care about.
The same logic runs through green bonds, sustainability-linked loans, and project-finance structures that reduce the cost of capital. Ørsted has been using green bonds since 2016, because offshore wind does not win by looking cheap on a spreadsheet at the factory gate. It wins by lowering the cost of financing the thing over decades. This is the less glamorous half of the energy transition, and the half that now matters more.
Good sites are becoming scarce assets
If hardware is less variable, location becomes more valuable. A strong site is a place with high output, low permitting pain, manageable environmental constraints, and a grid connection that does not eat the margin alive.
That sounds obvious until you look at how many projects were sold on the assumption that technology would bail them out. It often did. Now the board has changed. A better site can mean a higher capacity factor, which is simply the share of time a plant actually produces useful power compared with running flat out all year. It can also mean lower balance-of-plant costs, shorter cabling, fewer interconnection headaches, and less curtailment when the grid is saturated.
This is why the map matters again. A project in a mediocre location with cheap modules is no longer obviously better than a smaller project in a superior location with a stronger connection and a better buyer. The old logic treated all megawatts as equal. They are not. A megawatt that sits idle half the time is a far less convincing asset than a slightly more expensive one that keeps turning cash into electricity.
Utilisation is the new battleground
Higher utilisation is where a lot of the next value will be found, especially in hybrid projects that pair solar or wind with batteries. A battery does not make electricity cheaper by magic. It lets the project sell power when the grid wants it, not just when the sun or wind happens to show up. It also helps reduce curtailment, which is the polite industry word for being told to stop generating because the network cannot absorb any more.
Vistra’s Moss Landing facility in California, operating since 2020, is a useful reminder that storage is not a side dish. Once it sits beside generation or near a constrained grid node, it can change the shape of revenue completely. The same megawatt-hour is worth more if it arrives at the right hour.
Operators are also leaning harder on better maintenance, predictive analytics, and tighter asset management. Siemens Gamesa, like the rest of the turbine business, lives inside that pressure. Keeping a machine running, and catching faults before they turn into downtime, matters more when the hardware itself is no longer offering big cost relief. The machine has to earn its keep.
The new winners will be system builders
A lot of the old renewable marketing collapses here. The sector still talks as if falling panel prices are the main story. They are not. The real competition is moving toward finance, siting, grid access, operating discipline, and contract design.
The developers who understand that will do well. They will secure cheaper capital, choose better land, extract more output from each asset, and lock in power sales that lenders trust. The ones who still wait for the next miracle in module pricing will discover a brutal truth: a cheap component does not save a bad project.
