There are rough seas ahead for the U.S. wind industry. As the post-PTC era approaches, the industry will experience downward margin pressure, and wind farm owners will find it increasingly difficult to maximize profitability. Minimizing the Levelized Cost of Energy (LCOE) within operations without sacrificing long-term machine health will be a challenge for many. As an industry, success in the years ahead requires a clear-eyed accounting for all key inputs and considerations that go into wind farm operations.
Now in Part 2 of the frugal owner series, I want to outline costs and associated risks in more detail.
Let’s walk through the cost inputs I outlined last month:
How do you model your opex costs, 5, 10, and 20 years from now for new technology platforms across multiple suppliers?
The process of modeling failures and associated costs over 20-plus years is a challenge OEMs find themselves scratching their heads over from time to time. OEM’s such as Vestas spend hundreds of millions of dollars on failure testing within technology variations, and yet we still see this as one of the most significant risks in long term modeling. As you can see from the below chart, two different OEM wind farms of similar size and similar technologies were plotted against one another. The operational costs can vary by 40+% over time on these similar technologies due to variances in major component failure rates. You might wonder why? Aside from project specific climatic conditions, how well the site is maintained can materially affect the long-term health of the machines. As asset owners introduce more and more suppliers into each operating asset, thus incorporating multiple points of accountability, the long-term consequences can be material.
While I am not able to reveal our actual cost-basis, the cost figures in the chart represent current market prices for each scope identified. Our analysis proves that modeling forward costs can have a high degree of variance, and without a significant population of failure data, self-performers should be reviewing their in-house forward modeling with a severe level of scrutiny and making sure they include an appropriate risk factor to cover the uncertainty.
It’s also important to note that OEMs such as Vestas invest as much as $50MM annually into technology advancements and service-related IP. The cost projections in the table above do not include cost-out assumptions that we make based on our own internal investment in service-IP. Thus, without service contracts giving you price certainty, self-performers should be cautious when using the OEMs forward pricing curves as the basis for their in-house forward assumptions long term. This is a risk that is overlooked in modern day self-perform cost projections.
How much is your fixed cost allocations for administration, accounting, finance, etc. across all your vendors?
OEMs often find themselves in conversation across their internal teams regarding fixed capacity cost (i.e., Management, Finance, Accounting, HR, etc.), how it’s allocated, and what it means to the cost-basis when modeling customer assets. While organizational structures vary between OEMs, ISPs, and asset owners, their overall costs tend to be fairly similar. As you can see from the below table, as your fleet size grows, your fixed cost allocation becomes lower as a percent of the overall O&M costs. As a result, Vestas’ scale of operations can mean 7-8% of cost difference per project when compared to self-performing.
How much are you spending for labor? How much labor are you sharing / allocating on average per site?
The ability for an OEM to service wind farms with reduced labor costs is something difficult to match. A fundamental reason for this is that OEMs have a high density of project sites. This geographical advantage allows OEMs to site technicians strategically and reach more projects from one location rather than having disparate assets being serviced by different teams. The below maps show a general example of how nearby windfarms might share labor:
As a general rule, labor optimization is directly related to project size. More specifically, OEMs such as Vestas realize labor cost savings of +10% due to shared labor across operating sites. As the table and the maps together show, due to the massive Vestas service fleet, we expect that every site essentially acts like a 375MW site at a minimum no matter how small it is because chances are it’s near other sites.
(Vestas service sites spread across the U.S. by size.)
How much buying power do you have for consumables, minors, majors, and crane vendors compared to the OEM?
OEMs bring the scale of operations that give them cost-cutting abilities no asset owner can replicate on their own. In North America, the top three self-performers combined will have 40GW under their belts by 2025. To put things in perspective, that’s 20% less than the 50GW Vestas will have under contract. Currently, Vestas has 33GW under contract in North America while all self-performers combined have about 30GW.
You might be asking, “well, what’s the bottom-line impact this has on cost?”
Due to the multi-billion-dollar spend with suppliers to manufacture and operate wind turbines, OEMs such as Vestas provide an average savings of 15% to 20% on high volume orders for components. For lower volume and specialty orders, OEMs can save customers anywhere from 10-12% compared to buying directly from our suppliers.
The path forward
All this matters a great deal during a time when the money is going to soften for asset owners, and VPs of operations will need to find ways to keep costs down. Without this type of scrutiny, asset owners might find themselves with unachievable budgets unless they have appropriately accounted for future costs and associated risks.
(This is Part 2 of The Frugal Owner Series)