For most of the last two decades, procurement on a large capital project optimised against price. The market was assumed to be there; the question was what it would cost and who would win.
In 2026 that assumption no longer holds for a growing share of a project’s scope. The market is not there. Large power transformers are quoted at lead times of up to four years. The gas turbine slots are sold through 2028 and filling fast for 2030. The fabrication yards and engineering teams are committed. The contractors are declining to bid on work they judge too complex or too fast.
This is not a price problem with a procurement solution. It is a capacity problem, and it changes what sourcing is for: not selecting the cheapest capable supplier, but securing a position in a supply market that is allocating rather than competing.
The evidence is unusually clean, because much of it comes from the supply side’s own financial reporting.
Gas turbines are being sold as manufacturing slots before projects exist. GE Vernova’s second-quarter 2026 results, reported 22 July 2026, disclosed remaining performance obligations of US$176 billion and quarterly orders of US$24.2 billion, up 88% organically. The relevant line for a capital project is this one: Gas Power equipment backlog and slot reservation agreements grew from 100 GW to 116 GW, with at least 125 GW under contract expected by the end of 2026. Of that 116 GW, roughly 53 GW is firm equipment backlog and 63 GW is slot reservations — customers paying to hold a manufacturing position.
The pricing and availability trend inside those numbers is starker. GE Vernona reported first-half 2026 gas turbine pricing 10 to 20 points higher per kilowatt than in the fourth quarter of 2025, and roughly 10 GW of combined remaining availability across 2029 and 2030 – down from about 10 GW in 2029 alone a quarter earlier. Wood Mackenzie, in the same reporting, projects gas turbine prices reaching approximately US$600/kW by the end of 2027. Chief executive Scott Strazik was explicit about why customers were buying so far out: “We sold a lot of 2030 slots because the reality is we had a lot of customers looking at planning with EPC schedules and other dynamics needed the ’30 slot more than ’29.”
Siemens Energy tells the same story from a different book. Its third-quarter FY2026 release, 5 August 2026, reported an order backlog of €162 billion, record quarterly orders of €17.9 billion and a company-reported book-to-bill of 1.57. Its gas turbine backlog stood at roughly 60 GW at the end of the second quarter of FY2026, booked through fiscal 2028, with the company’s head of investor relations telling a pre-close call on 29 June 2026 that 2029 and 2030 slots were “filling up very quickly” – and Siemens Energy raised its view of the global gas turbine market to 110–120 GW per year, from 90–100 GW. Mitsubishi’s large-frame gas turbine backlog reached 35 GW, up from 23 GW a year earlier, with orders taken in the first quarter of FY2026 delivering in 2028–2030.
Read those three disclosures together. The world’s three large gas turbine manufacturers are all reporting multi-year sold-out positions, rising prices, and a commercial model in which capacity is reserved years ahead of a defined project.
Electrical equipment is worse, and it is official. The International Energy Agency’s Building the Future Transmission Grid analysis (25 February 2025) documented procurement lead times of two to three years for cables, up to four years for power transformers, and beyond five years for DC cables, with cable costs nearly doubling since 2019 and power transformer prices up around 75%. The IEA’s Electricity 2026 update reports that prices for key grid components have nearly doubled over the past five years, and that over 2,500 GW of renewable, large-load and storage projects now sit stalled in grid queues worldwide. The earlier report had already counted 1,650 GW of solar and wind projects at advanced stages of development waiting on connection.
Survey data fills in the equipment detail. Wood Mackenzie’s Q2 2025 transformer survey, reported by POWER magazine in January 2026, put average lead times at 128 weeks for power transformers, 144 weeks for generator step-up transformers and 44 weeks for high-voltage switchgear – with power transformer prices up 77% and some distribution transformer classes up 95% since 2019. Demand over 2019–2025 rose 119% for power transformers and 274% for GSU units, against a 2025 power transformer supply shortfall of around 30%.
The mechanism is stated most clearly by the IEA itself: grid infrastructure takes five to fifteen years from planning to completion, while a data center takes one to three years. Both draw on the same transformer, switchgear, cable and electrical-contracting queues – and the shorter-cycle buyer gets to the front of the line, repeatedly.
That competition is now visible in supplier order books. GE Vernona reported more than US$5 billion of data center orders year-to-date in Electrification by the second quarter of 2026 – more than double its full-year 2025 total. Turner & Townsend’s Global Construction Market Intelligence 2026 finds investment shifting toward data centers, advanced manufacturing and AI-related infrastructure while commercial and residential demand softens and identifies data centers as the sector where contractors report the highest profit margins, cited by 27% of respondents.
For a refinery, LNG train or mine competing for the same switchgear, the same E&I contractors and the same craft labour, that margin differential is the whole problem. Your project is bidding against a buyer with a shorter schedule, a higher tolerance for equipment premiums, and a better margin to offer the contractor.
Turner & Townsend’s 2026 edition covers 112 markets across 44 countries. Its labor findings are severe: approximately 77.7% of markets report labor shortages or severe labor shortages, 16.1% report balanced conditions, and fewer than 6% report any form of surplus capacity. Regionally, 100% of Australia and New Zealand markets are in shortage, Europe exceeds 93% and North America sits around 79%. Over two-thirds of markets report shortages of skilled green labor – which the report identifies as a key limitation for energy infrastructure projects seeking better sustainability credentials. Global construction cost inflation ran 4.2% in 2025 and is forecast at 4.5% for 2026.
The behavioural consequence is the part procurement should read twice. The report finds that contractors are becoming more selective, particularly on complex and fast-track projects, describes reduced appetite from bidders and increased subcontractor selectivity, and notes margins and preliminaries rising to reflect greater caution. Its own recommendation is procurement strategy, not commercial strategy: early engagement and collaborative procurement to secure capacity.
On craft labour, Associated Builders and Contractors’ 15 January 2026 model estimates the US construction industry must attract 349,000 net new workers in 2026 and 456,000 in 2027, derived by converting inflation-adjusted construction spending into labour demand at roughly 3,450 jobs per US$1 billion of additional spending. Chief economist Anirban Basu’s framing matters more than the headline: a majority of 2026 demand is attributable to retirement rather than growth, with approximately one in five electricians over 55. That is a structural withdrawal of capacity, not a cyclical shortage that a rate cut fixes.
The clearest statement that capacity, not capital, is the binding constraint came from Rystad Energy on 20 May 2026. Analysing a Gulf reconstruction requirement of US$34–58 billion, Rystad concluded that “capital availability is not the primary constraint; instead, access to equipment, contractors and logistics is emerging as the key limiting factor,” and found that long-lead equipment procurement defines recovery timelines.
It then named a displaced megaproject: Qatar’s North Field expansion faces potential delay because repair and expansion work compete for “similar pools of engineering teams, fabrication yards and site crews.” Senior analyst Karan Satwani put it directly: “The $58 billion bill is the headline, but the knock-on effects on energy investment timelines globally may prove just as significant.”
There is precedent for the outcome. In February 2025, ENGIE Flexible Generation withdrew two Texas peaking plant projects, with managing director Eric De Caluwe citing “equipment procurement constraints, among other factors,” as making the statutory schedule unachievable – at a point when, as the same reporting put it, the largest manufacturers were telling buyers they should be talking to their OEMs as much as seven or eight years out. A well-capitalized developer with a viable project abandoned it because the supply market said no.
Worth noting what the claims literature does not yet show. HKA’s CRUX Insight Eighth Annual Report (14 November 2025) analyses more than 2,200 projects across 114 countries with combined capital expenditure of US$2.433 trillion and ranks post-2020 causes as change in scope 28%, COVID-19 24%, design issues 15% and payment disputes 14%. It does not isolate supplier or vendor capacity, equipment delivery or subcontractor capability as a ranked cause. The industry’s largest claims-causation dataset has not yet caught up with what the OEM order books are demonstrating – which means a project relying on dispute statistics to size this risk will underestimate it.
KPMG’s Global Construction Survey 2025/2026, published March 2026 from 375 industry leaders, contains the single most useful pairing in this article:
The same survey finds only 20% consider collaborative contracting standard practice today, against 56% who expect it to be standard within five years; 70% plan to continue relying on contingent workers, with in-house employees accounting for just 46% of the workforce; and one in three ranking material and equipment costs and financing constraints among their biggest challenges. One caveat on vintage: KPMG’s fieldwork ran from January to March 2025, though the report published in March 2026.
The capacity to act on this is also tightening. The Hackett Group’s 2026 Procurement Agenda and Key Issues study (17 March 2026) projects procurement workloads rising 8% in 2026 while both headcount and operating budgets are expected to decline. AI-enabled technology entered procurement’s top three priorities for the first time – 43% are actively pursuing deployment, though only 12% at large scale – and current deployment concentrates in contract management, market intelligence and spend analytics.
Deloitte’s 2025 Global CPO Survey (250+ CPOs across 40 countries) shows which risk responses executives actually rate as effective: maintaining alternative sources 74%, supply chain visibility 64%, and supplier information sharing and collaboration 61%. All three are supplier-data capabilities before they are commercial ones.
There is no published benchmark for the share of organisations performing formal supplier capacity assessment. Given 17% have supply chain mapping tools at all, the honest inference is that it is a small number.
Three named, dated examples define the emerging practice.
Pre-FID long-lead ordering. On 22 December 2025, Commonwealth LNG authorized Technip Energies to issue purchase orders for long-lead equipment for its 9.5 Mtpa Cameron Parish facility – valued at €250–500 million – before its final investment decision, then expected in the first quarter of 2026. The scope covered six mixed-refrigerant compressors with Baker Hughes LM9000 gas turbines, six Honeywell main cryogenic heat exchangers and four Solar Turbines Titan 350 gas turbine-generators. The stated rationale was the long lead times required to support an accelerated modular construction schedule. Capital was committed to equipment before the project was formally sanctioned, because the queue would not wait for governance.
Buying contractor and factory capacity as a strategic act. On 5 November 2025, AEP and Quanta Services announced two agreements: a Cooperation and Commitment Agreement covering joint design, engineering, procurement and construction of 765 kV and high-voltage transmission facilities, and a Development Services Agreement to expand domestic manufacturing capacity for extra-high-voltage transformers and circuit breakers – framed within AEP’s US$72 billion 2026–2030 capital plan. Quanta’s chief executive described it as delivering “the labor certainty and capacity needed.” An owner underwrote craft labor and OEM factory capacity years ahead of need.
Slot reservation as standard commercial practice. The 63 GW of GE Vernova slot reservations, and Siemens Energy’s reporting of “orders plus reservation agreements” as a combined commitment metric, mean paying to hold a build position is no longer exotic. It is how the power equipment market now clears.
A caution against over-claiming: no published outcome study measures the cost or schedule benefit of early vendor engagement or capacity reservation. Turner & Townsend’s recommendation is expert judgement, and KPMG’s collaborative-contracting numbers measure adoption rather than results. The case for acting early rests on the arithmetic of queue position, not on a proven effect size.
Five shifts, each of which is a supplier-data requirement before it is a commercial one.
Bring supply market capacity analysis forward to FEED. The question at FEED is not what a compressor train costs but when the next available slot is, and whether the schedule the project is about to commit to is physically obtainable. Every long-lead item needs a market-availability position, dated, attributable and refreshed – not a budget price from a two-year-old estimate.
Score capacity, not just capability, in prequalification. A supplier that is technically excellent and fully committed through 2029 is not a qualified bidder for a 2028 delivery. Current backlog, committed capacity, expansion plans and the buyer’s likely queue priority are prequalification attributes. Almost no supplier registration process captures them.
Track sub-tier capacity, because that is where the constraint usually sits. The queue is rarely at your Tier 1. It is at the transformer works, the forging house, the specialist coating shop or the single certified fabricator inside the contractor’s own supply chain. With only 17% of construction leaders using supply chain mapping tools, most projects find this out during expediting.
Reserve early where reservation is possible, and evidence what you reserved. Slot reservations, capacity agreements and pre-FID long-lead orders create commitments that need to be held as structured records against the package, the schedule milestone and the approval that authorized them – not as an email thread.
Treat contractor selectivity as a live sourcing risk. When more than three-quarters of markets report labor shortages and contractors are choosing which complex work to bid, bid coverage is no longer a given. Early engagement, realistic risk allocation and a credible project reputation become procurement deliverables. Which contractors declined to bid, and why, is data worth capturing – most projects never record it.
Each of those shifts assumes something most capital projects do not have: a current, structured, shared view of the supply base – capacity, capability, commitments, sub-tier relationships and compliance status – that engineering, construction and procurement all read from and write to.
Dharas provides that layer. Suppliers self-register and maintain their own profiles, so capability, capacity and compliance data arrives as governed structured records from principal contractors down to Tier-N vendors, rather than as attachments in someone’s inbox. The supply base is searchable and filterable by status, region and attribute, so a sourcing team can answer “who else can actually build this, and are they available” from a record rather than a phone call. Two-way notifications let the project issue expressions of interest and capacity enquiries and receive responses in-platform, on a defined cadence. Dashboards show the supply base and its compliance status in real time, and structured data exports out at any point with auditable lineage.
None of that manufactures a transformer. It does mean that when the schedule depends on knowing which suppliers have capacity, the project has an answer it can defend – and a record of what was reserved, when, and by whom.
Three gas turbine manufacturers are reporting sold-out positions into the end of the decade. The IEA has power transformers at lead times of up to four years and grid component prices roughly doubled in five. Turner & Townsend has 77.7% of construction markets reporting labor shortages, with under 6% reporting any surplus. Rystad, looking at a US$34–58 billion reconstruction bill, concluded the constraint is equipment, contractors and logistics – not capital. And three in four construction leaders agree the risk is elevated while 17% have the tools to map it.
A procurement head commissioning a new capital project in this market has a different job than the one the role description describes. The task is no longer to run a competition and select the best offer. It is to establish, early and with evidence, whether the supply market can build what the schedule promises – and to secure a position in it before someone with a shorter cycle and a better margin does.
The projects that fail in the next three years will mostly not fail because they paid too much. They will fail because they arrived at the queue too late, and had no data telling them the queue existed.
Local content law now binds subcontractors directly and penalizes the operator. What DRC, Cameroon, Senegal...
READ ARTICLE →
Industrial organizations authorize an average of 77 third parties into their OT environment; a quarter...
READ ARTICLE →
A steel importer settled at $19 million in May 2026 for declaring Chinese and Turkish...
READ ARTICLE →