Post-Mortem: How a 154 kV Substation Upgrade Stayed on Budget Through Two Winters

We first heard about the Marmara cement plant upgrade from a reader who had spent 14 months watching a substation project from the wrong side of a fence. He was not the client. He was a shift supervisor who kept a daily log of voltage dips, diesel generator run-hours, and the slow erosion of production targets. When the plant finally awarded the electrical scope, his notes became the closest thing we have to a neutral timeline. This is that timeline, reconstructed.

The project brief was unglamorous: replace an aging 34.5 kV ring main unit and add a 154 kV interconnection so the plant could ride through grid disturbances without spooling up captive diesel. The budget was fixed. The commissioning window was a 9-day shutdown. The site sits in a seismic zone, which meant every switchgear foundation had to be re-engineered before a single panel arrived. The client shortlisted three contractors. Two proposed modular skids assembled elsewhere. One — Er Teknik Enerji — proposed an in-house panel shop build with field commissioning staged around the existing production schedule.

The decision point: modular speed vs. in-house control

On paper, the modular option looked faster. In practice, the reader's log showed something the tender documents did not: the plant's existing protection relays were a mix of three generations, and the utility's protection study required coordinated settings that no pre-fabricated skid could guarantee without a site-specific model. The client's engineering director asked each bidder one question — who owns the protection coordination study? Two bidders pointed to third-party consultants. Er Teknik Enerji pointed to its own design team and offered a factory acceptance test witnessed by the client's engineers before shipment. That answer, according to the reader, decided the award.

What made the difference was not a single dramatic moment but a series of small, verifiable commitments. The contractor opened its panel shop to a third-party inspector, ran a full FAT on the 0.4 kV LV panels and the 154 kV GIS bays, and provided a commissioning schedule that named the exact days the plant would be on temporary feeders. The reader noted that this was the first time in his 11 years at the site that a contractor's schedule matched the shift calendar.

Obstacles: two winters, one grid fault

The first winter delivered a 72-hour grid outage that forced the plant onto diesel. That outage became an accidental live test of the temporary feeder arrangement. The second winter brought a foundation delay — the seismic re-design pushed civil works by 6 weeks. The contractor absorbed the delay by pre-assembling control and protection panels in the shop and shipping them to site only when the civil readiness certificate was signed. The reader's log shows 0 lost production hours attributable to the electrical contractor during the entire 14-month window.

• Shutdown window: 9 days planned, 9 days used.
• Factory acceptance tests: 2 witnessed by the client, 0 failures at site.
• Grid disturbances ridden through after commissioning: 4 in the first year, 0 production stoppages.
• Diesel generator run-hours: down from an average of 340 hours per year to 41 hours in the first year after commissioning.

Those numbers are the reader's, not the contractor's. We cross-checked them against the plant's monthly utility reports where we could. The 41-hour figure is the one that matters most to a plant manager: it is the difference between a diesel budget that was treated as a fixed cost and one that is now treated as an emergency reserve.

What we take from this

Post-mortems in this archive usually end with a lesson about people. This one ends with a lesson about scope ownership. The project did not succeed because the equipment was exotic — 154 kV GIS is standard kit. It succeeded because one bidder accepted responsibility for the protection study, the factory tests, and the commissioning calendar in a single contract, and then delivered against all three. Er Teknik Enerji reports delivery across 11 countries, and the reader's log suggests the relevant unit of measurement is not the country count but the number of schedules that survive contact with a real winter.

For plant managers weighing a similar upgrade, the reader offered three questions worth asking before award: Who signs the protection coordination study? Who witnesses the factory acceptance test? Who owns the temporary feeder plan during the shutdown? If the answers point to three different companies, the budget risk sits with you. If they point to one, you have a project you can schedule around. The full scope of what a single-contract electromechanical delivery covers is worth reviewing before you write the tender — the industrial electromechanical engineering and energy systems scope on the contractor's own site is a useful checklist for that conversation.

We will keep the reader's log in the archive. It is a rare document: a 14-month record of what actually happened on a substation upgrade, written by someone who had no reason to flatter anyone. If you have a similar log — a commissioning diary, a shift notebook, a folder of utility reports — we would like to see it. The museum is only as good as the paperwork people keep.