How to Choose a Wind Turbine for Wholesale: A 7-Step Procurement Checklist

Who this checklist is for

If your team is sourcing wind turbines for a utility-scale or distributed project this year, this is the process we actually run before anyone signs anything. Not the version in a conference deck.

I'm a procurement manager at a renewable energy developer — around 900 people, mostly project development and O&M. I've been running our equipment and services budget for eight years now. Roughly $140 million in cumulative spend over that period, about a third of it turbines and balance-of-plant. Every order goes into our cost tracking system, and we review it quarterly.

Seven steps. Six weeks if you move fast, ten if you don't.

One assumption up front: you're not building turbines yourself. You're choosing between manufacturers. That's a different problem, and the failure modes are different too.

Step 1 — Start from the grid side, not the product catalog

Most teams do this backwards. They call an OEM, get three spec sheets, and pick the one with the best numbers. That path almost always loops back.

The right order: settle the interconnection constraints first. Is it 110 kV or 220 kV? What reactive power compensation is required? What's the short-circuit capacity at the point of interconnection? What's the low-voltage ride-through requirement?

Those constraints dictate which turbine classes you can even quote. The same 6 MW platform can carry a seven-figure cost difference in tower and collector system design depending on the connection voltage. We had a project in 2023 where we moved too fast — signed the technical spec, then discovered the POI capacity couldn't support the model we'd chosen. We dropped a class. Six weeks of schedule, gone.

Deliverable for this step: a one-page interconnection constraint sheet. One page. Then call OEMs.

Step 2 — Translate IEC 61400 classifications into procurement language

The IEC 61400 series governs turbine design standards. IEC 61400-1 covers onshore design requirements, IEC 61400-3 covers offshore. Every OEM engineer knows these documents cold. Most procurement teams don't, and that gap is where money leaks.

You don't need to read the standards. You need to convert them into contract language.

So instead of writing "complies with IEC 61400-1" in the technical specification, write the four numbers that actually move the price:

• Turbulence class (IEC Class A, B, or C)
• Wind shear assumption
• Extreme wind speed design value
• Design service life

Change any of those and quotes can swing 15%.

I learned this the hard way. We negotiated on a "standard IEC 61400-1 platform." Then the OEM pointed out their default was Class B turbulence, and our site measured Class A. Upgrading the configuration cost us extra, mid-negotiation.

Since then, page one of our technical spec is those four fields. No fields, no bid invitation.

Step 3 — Force every quote into the same cost model

This is the instinct part. But in wind, cost normalization is more dangerous than it looks, because what's included in a per-MW quote varies wildly between vendors.

Common inclusions: tower, foundation embedments, pitch system, main control system, SCADA, spare parts package, installation supervision, commissioning, and warranty-period O&M. Some OEMs quote all of it. Some quote almost none of it.

We run every bid through a six-line TCO model: equipment delivered, site delivery and erection, grid commissioning, warranty-period O&M, post-warranty service contract, spares and consumables. Discounted over twenty years.

When you do this, the lowest per-MW bid often doesn't place in the top two.

Last quarter we had a case where Bid A came in 6% under Bid B per megawatt. After TCO normalization, B won — A hadn't included installation supervision or warranty-period consumables.

I didn't go back and squeeze A on price. On a twenty-year asset, consistency matters more than a few points on the quote sheet. And that's not a philosophical position, it's an accounting one. An underperforming fleet doesn't just hurt the P&L — it hurts your next PPA negotiation, your lender's confidence in your delivery record, and your standing with the local grid operator. Saving on output is a bill that comes due in client perception.

We went with B. That bid happened to include a Goldwind platform, but the model picked it, not the brand.

Step 4 — Verify manufacturing and delivery, not lab data

OEMs show you lab data and reference project performance. Both matter. Neither is what you're contracting for.

What you're buying is: how many units, delivered where, on what cadence.

Three questions to ask:

1. How many units of this specific platform were delivered globally in the last twelve months? Not cumulative — last twelve months.
2. Where are the towers and blades manufactured, and is there a capacity conflict? If three other projects need the same blade plant next year, where do you sit in the queue?
3. Who owns the installation resource — the OEM, or a third party? How is the third party's schedule locked?

I've watched a signed contract slip four months because blade capacity got redirected to a larger project. The financial carry on those four months dwarfed the original price spread.

OEMs with assembly footprints outside their home market tend to hold delivery schedules better, especially on European or Americas projects — Goldwind being one of the more visible examples. But you still verify project by project, because the same brand's delivery performance varies a lot by region. That's why I track goldwind news on manufacturing footprint, not to pick a brand, but to anticipate delivery risk.

Step 5 — Layer the warranty. Don't accept a single availability number.

Availability is a word that can be engineered in a contract to mean almost anything.

A 97% figure might be calculated on turbine-hours, excluding curtailment and planned maintenance. Or it might be calculated on available generation hours, including planned maintenance. Same number, two percentage points of real difference.

We layer ours into three tiers.

Tier one — definition. The calculation basis, the measurement window, and every exclusion, written out. No "standard industry practice."

Tier two — remedy. What happens when the threshold is missed. Compensation formula. Cap.

Tier three — exit. If a platform misses two consecutive measurement periods, is there a replacement or swap mechanism?

Tier three is the one most buyers skip. It's also the one that actually protects you.

And to be clear: don't accept a "100% guaranteed availability" clause from anyone. It's not physically achievable, and a contract term that can't be met is just noise.

Step 6 — Lock local content and spares logistics into the contract

This is the step that gets dropped. You're five steps in, the team is tired, the contract is almost signed.

Local content and spares logistics are exactly where post-commissioning problems live.

Local content isn't just a policy question — US IRA-related provisions, various European localization requirements. It directly drives your O&M cost. A component shipped from overseas averages two weeks by sea plus customs. Peak season, longer. Over two years, generation lost to spare-part waiting can exceed the entire equipment price delta you negotiated.

So contract three things: local stock requirements for critical spares, emergency dispatch response time, and a compensation mechanism when response time is missed.

Our current standard: main control boards, pitch motors, and gearbox bearings must be stocked regionally with a 48-hour response commitment. Since we wrote that in, downtime hours from spare-part delays dropped by roughly 60%.

Step 7 — Run the bad-weather drill

Last step, and the one I see least often across the industry: before signature, walk the worst case.

Two hours. Technical, O&M, finance, legal in the room. Four questions, in order.

1. If the installation window slips three weeks due to sustained high winds, who carries what?
2. If the interconnection is delayed two months, who pays for storage and insurance?
3. If the first-year fleet shows a batch quality issue, how is lost generation calculated?
4. If the OEM itself has a delivery problem, what's your fallback?

Unasked, the risk defaults to you.

We ran this drill on a project last year and found the force majeure definition was broad enough to cover the supplier's own capacity problems. We renegotiated. The revised clause got used on a later project.

Where this usually goes wrong

• Treating the quote as the cost. It's the opening position, not the number.
• Treating IEC compliance as a quality guarantee. It's a floor, not a ceiling.
• Optimizing year one instead of year twenty. The asset outlives everyone who approved it.
• Writing warranty terms loosely and fixing them later. Later doesn't happen.
• Skipping the spares logistics test. A supply chain on paper and a supply chain in Rotterdam are different things.
• Accepting verbal commitments. Delivery dates, upgrades, pricing — email or contract appendix. Nothing else.

None of these seven steps is complicated. Every one of them has a reason to skip — tight timeline, tight budget, an OEM pushing to close. In my experience, every skipped step costs double to fix later.

If you're running wholesale wind turbine sourcing this year, print the list. Tick each line. The one you can't tick is your project's actual risk.