The TMI-1 restart and the Susquehanna campus proved the commercial template: nuclear plus data center is now a deal category, and every hyperscaler is running the same site search at the same time. The winner is rarely the best site; it is whoever answers the feasibility question first. Meridian Point Energy operates the two-unit Harbor Bluff station, 2,150 MW, when the inquiry lands: an AI campus, 800 MW ultimate build, first 200 MW energized inside 24 months, 24/7 carbon-free preferred. The old answer is a 9-to-12-month manual feasibility study, and the deal walks while it runs. This is the other answer: a feasibility twin, a constraint sweep, three configurations scored on the same physics, and a bankable term sheet in six weeks: 800 MW, 20 years, refueling outages priced on day one.
| Without | With GridCORTEX | Δ |
|---|
| Without | With GridCORTEX | Δ |
|---|
The Term Sheet follows Meridian Point Energy, a fictional operator of the two-reactor Harbor Bluff nuclear station (2 × 1,075 megawatts, 2,150 MW total, enough power for roughly two million homes). The clock starts on Day 0, when an inquiry lands from a hyperscaler, one of the giant cloud-computing companies. It wants an AI campus: 800 MW at full build, with the first 200 MW running inside 24 months, powered around the clock by carbon-free energy if possible. The same inquiry is running at three other sites, and the buyer will commit to whoever answers credibly first. The conventional answer is a feasibility study done one piece at a time over 9 to 12 months; deals die waiting for it. The deal team instead accepts a 6-week window. The feasibility question breaks into 214 specific checks across the grid, the plant, regulation, land, and water. In 7 days the software assembles a feasibility twin, a working computer model of the whole site used to test the deal safely before committing to it. The twin holds every grid connection within 25 miles (42 network junctions, 9 substations, and both major high-voltage corridors), the waiting line of 3 earlier grid-connection requests ahead of this one, both reactors' operating limits and the power the plant consumes to run itself, the security and emergency-planning boundaries required by the Nuclear Regulatory Commission (the federal nuclear safety regulator), and two candidate parcels of land: north (310 acres) and southeast (540 acres).
Week 2 runs all 214 checks at once, and it front-loads the bad news. The north parcel is ruled out: 60 percent of its buildable area sits inside the federally required nuclear security setback. Serving the full 800 MW by plugging the campus directly into the plant, behind the utility meter, would strain the power the plant needs for its own equipment and trigger a grid-reliability review. And the transmission lines can carry 430 MW to the campus through the grid today, rising above 900 MW after a corridor upgrade finishes in month 30. The good news lands too. Water clears; the third phase needs an 8-month permit amendment that blocks nothing. The plant's refueling shutdowns, the planned outages every 18 months when a reactor loads fresh fuel, clear because they can be priced into the contract on day one rather than discovered as a surprise later. And the plant's existing grid-connection rights can carry the first block of grid-delivered power with no new wait in the connection queue. The sweep ends: 214 checks, 6 serious constraints, 4 cleared or fenced in, 2 left as design choices. Overnight, the software scores 3 deal structures against 40 what-if cases each, varying prices, schedules, and demand. Decision Point 1 asks the deal team to choose. Option A plugs the campus straight into the plant: first power in month 20, but the most legally untested path. Option B serves it entirely through the grid: the cleanest path, but first power in month 27, missing the 24-month requirement. Option C, the recommended hybrid, starts with a 200 MW direct connection in month 22 and grows through the grid to 800 MW by month 34.
With Option C locked, the term sheet, the short document of deal terms both sides negotiate from, assembles straight out of the twin. Its spine: a 20-year contract; capacity stepping up in blocks of 200, then 450, then 800 MW on dates the grid can actually meet, with automatic adjustments if construction slips; carbon-free power matched to the campus's usage hour by hour, at 91 percent to start and 96 percent once phase 2 adds more sources; refueling shutdowns handled as pre-priced, capped purchases of grid power; and a price band of $78 to $88 per megawatt-hour (the standard unit for pricing bulk electricity), which clears the buyer's economics in 34 of the 40 what-if cases. Decision Point 2 asks the team to release it inside the window. Approved, the package goes out in week 6, day 1: the term sheet plus a 240-page technical annex backing every number, then a board package. By day 40 the buyer asks for exclusivity, agreeing to negotiate with this site alone, while its other three candidates are still writing consultant contracts. Left unapproved, the simulation ends the other way: no configuration, no date, and a buyer that stops waiting in month five.
The failure is the sequence, not the people. Grid study, then land, then regulatory, then commercial: roughly 4,100 engineering hours across 5 outside firms over 9 to 12 months, all spent evaluating a single configuration priced by comparison with other deals. The fatal flaw, the nuclear security setback covering the north parcel, surfaces in month 7, after most of the money is spent. At month 5 the buyer stops waiting and moves its capital to the site that answered first. The study finishes anyway, describing a deal that no longer exists: $0 contracted, 800 MW and 20 years of revenue gone, and a binder on a shelf that helps nobody with the next inquiry.
The feasibility twin runs every workstream at the same time on one shared model, so the fatal flaw shows up in week 2, before a dollar goes to the wrong parcel. Three deal structures are scored on identical physics instead of one structure priced by comparison. The term sheet is then drafted so that every number traces back to the model: capacity steps on dates the grid can really meet, carbon-free percentages computed from the plant's actual output including outages, refueling shutdowns priced on day one. Total effort: about 380 engineering hours and a bankable answer in 6 weeks, an answer solid enough for boards and lenders to rely on, followed by exclusivity. Both decision points are human gates: the deal team, the lawyers, and the board approve every structure. The software drafts; people decide.
| KPI | Without GridCORTEX | With GridCORTEX | Delta |
|---|---|---|---|
| Elapsed time to answerhow long the buyer waits for an answer it can take to its own board | 9 to 12 months | 6 weeks | 7x to 10x faster |
| Engineering hourspaid expert hours consumed producing the answer | ~4,100 across 5 firms | ~380, aided by the twin | the team keeps its day jobs |
| Configurations evaluatedhow many possible deal structures were actually tested | 1, one step at a time | 3 structures × 40 what-if cases | all scored on the same physics |
| Confidence basiswhat the numbers in the offer actually rest on | spreadsheets + comparisons to other deals | a live model of the grid and the plant | every number traces |
| Fatal-flaw discoverywhen the deal-killing problem, the nuclear security setback, comes to light | month 7, after the spend | week 2, before the spend | north parcel out early |
| Reusabilitywhat is left over to answer the next buyer's inquiry | binder on a shelf | living model, ready for the next ask | the second answer is faster |
| Deal retentionwhether the buyer is still at the table when the answer arrives | walks by month 5 | exclusivity in week 6 | the load stays |
| First powerwhen the first 200 MW of the campus actually turns on | month 30+, if signed at all | month 22, phase 1 | the 24-month ask is met |
| PPA structurethe power purchase agreement, the long-term contract to buy power at agreed prices | generic template | hourly carbon-free matching + phased blocks | bankable, board-ready |
| Outage coordinationhow the plant's planned refueling shutdowns are handled in the contract | negotiated after signing | priced in on day one | no year-3 surprise |
| Price discoveryhow the offered price was arrived at | a single point estimate | a band tested against market prices | 34 of 40 cases clear |
| Contracted valuerevenue actually locked in by the process | $0 | a 20-year contract for 800 MW | the franchise grows |
This is a desk analysis and it moves almost no field exposure, so we will name the small real effects instead of inventing large ones. Nothing it produces is a licensing basis change, and any actual arrangement goes through your licensing review and your own NRC engagement before it is anything. The one genuine mechanism is that screening a parcel out on paper removes the site visits, switchyard walkdowns, and escorted entries that would otherwise be spent proving the same thing in person.
Counted in units you already track:
Inquiry response hours come back to strategy, transmission planning, licensing, and security staff, all of whom absorb this work on top of their existing assignments.
The numbers we need from you to run that formula:
| Cost driver | How it is calculated, from a rate you supply |
|---|---|
| Internal labor per inquiry | hours avoided per inquiry x your loaded rate for each contributing group x inquiries per year |
| Outside advisory and legal | advisor and legal hours avoided on first pass feasibility and term structure x your contracted rates, with counsel retained for anything binding |
| Wasted effort on non starters | inquiries screened out earlier x hours that would have been spent before the constraint surfaced x your loaded rates |
| Response speed | the value of answering in weeks rather than months, which only you can bound, from your own view of the contracts you compete for and the margin on them |
You pay for the scoped engagement that builds and runs this, for integration into your transmission planning data, site and security records, and document system, and for the staff time of the same experts whose hours you are trying to save, especially in the first inquiry. Legal, licensing, and commercial review everything before it leaves the building, and GridCORTEX sends nothing to a counterparty.
Payback is calculated on internal and advisor hours per inquiry across your annual inquiry volume, including the ones that do not proceed, because those hours are spent either way. Winning a contract is the reason to do it and it is not a number we will model for you.
A drafting and analysis assistant has no direct safety benefit and we will say so plainly. The indirect link is that a tariff without real curtailment and ramp obligations leaves operations to solve a large load problem in the moment, and in the moment solutions are the ones done under time pressure in the field.
Counted in units you already track:
Rate analyst and regulatory staff hours come back, and planning stops being interrupted for informal opinions on contract language.
The numbers we need from you to run that formula:
| Cost driver | How it is calculated, from a rate you supply |
|---|---|
| Rate and regulatory labor | analyst and regulatory staff hours avoided x your loaded rates |
| Outside counsel and consultant fees | advisor hours avoided on rate modeling and term sheet drafting x your contracted rates |
| Deal cycle time | negotiation weeks removed x your own carrying cost per week, plus the revenue start date moved forward x your expected monthly revenue from the agreement |
| Cross subsidy exposure | the rate base impact of terms you would have agreed without a modeled revenue and cost allocation, priced with your own cost of service allocation, which is exactly the number a rate case will test |
| Avoided wires acceleration | the capital you do not have to pull forward because the agreement's flexibility terms hold peak inside the existing plan, at your own project cost estimate and your carrying cost per year |
You pay for the assistant, for connecting it to your cost of service model, your load forecast, and your network model, and for legal and regulatory review of everything it drafts, because drafts do not leave the building without your people. The connection to your cost of service model is the integration that determines whether this is useful or decorative.
Payback is normally led by analyst and outside advisor hours plus deal cycle time, both of which you can audit from timesheets and invoices. Avoided cross subsidy is the strategically largest item and the one you will only prove in a rate case, so keep it out of the base case.
This is desk work and it puts nobody in the field. The only genuine safety mechanism is that sites which will not survive screening no longer get reconnaissance trips, preliminary field survey campaigns, and met tower mobilizations before somebody notices they were never viable.
Counted in units you already track:
Screening hours come back to the development team, transmission planning, and licensing, and the sites that reach a paid study are fewer and better chosen.
The numbers we need from you to run that formula:
| Cost driver | How it is calculated, from a rate you supply |
|---|---|
| Consultant screening scope | screening study hours you no longer buy x your consultant billing rate, plus the study fee for each site you eliminate internally before paying for it |
| Internal labor | development, transmission planning, and licensing hours avoided x your loaded rates for each |
| Site control carrying cost | option, lease, or land holding cost per month x months of holding avoided on sites that early screening eliminates |
| Interconnection queue | study deposit and queue application cost per site x applications you do not file on sites that screen out, plus whatever your tariff makes non refundable |
You pay for the scoped engagement that builds and runs this, for the data acquisition and integration to bring geospatial, seismic, hydrologic, meteorological, and interconnection datasets together at the resolution a screen needs, and for your own engineering, licensing, and transmission staff to review the output. You will still buy a consultant siting study for the sites that survive. This is not a replacement for that study and should not be sold internally as one.
Payback is dominated by consultant scope you do not buy and by development capital you do not sink into sites that were never viable. Both are auditable against your own past study invoices, so build the case there and leave decision quality out of the arithmetic.
A feasibility report has no direct safety benefit. The honest indirect mechanism is specific to co-location: a deal agreed before the plant side topology is understood produces late scope changes at an operating plant, and late scope at a running plant means work forced into an outage window that was already fully booked.
Counted in units you already track:
Hours come back to four departments at once, and the corporate development director stops being a document assembler.
The numbers we need from you to run that formula:
| Cost driver | How it is calculated, from a rate you supply |
|---|---|
| Cross department staff time | department hours avoided x your loaded rates for planning, plant engineering, regulatory, and commercial staff |
| Outside advisor fees | consultant and outside counsel hours avoided on feasibility scoping x your contracted rates |
| Deal capture | your own expected annual revenue from the deal x the probability you assign to your report arriving inside the developer's decision window being the reason you win it, a probability you set, not us |
| Bad sites screened early | your own average spend on a co-location deal that dies in late diligence x the number of such deals per year an early screen would have stopped, a count your team judges |
| Reinforcement right sizing | the difference between the reinforcement scope a coarse screen assumes and the scope the detailed model supports, at your own unit costs |
You pay for the feasibility service, for the work to load your plant models, network model, and interconnection data into it, and for commercial, legal, and engineering review time on every draft, because nothing goes to a developer without your people signing it. The internal review discipline is the cost that people forget to budget.
Payback is carried by staff and advisor hours across four departments, which you can audit. Deal capture is the number that dwarfs everything else and the one you should present as upside, because you cannot prove the counterfactual.
This is a modeling service and it has no direct field safety benefit. The indirect mechanism is credible: a large critical load that does not ride through a disturbance produces an emergency restoration, and emergency restorations put crews into energized work and short notice switching at speed.
Counted in units you already track:
Planning engineer hours and negotiation hours come back, and the strategy analyst stops hand carrying findings between engineering and legal.
The numbers we need from you to run that formula:
| Cost driver | How it is calculated, from a rate you supply |
|---|---|
| Planning and consulting engineering | scenario hours avoided x your loaded engineer rate, plus outside study fees avoided at your contracted rate |
| Contract exposure | your own service level or availability commitment cost per event x the number of events the model shows the current design cannot ride through, a count you decide is credible |
| Design rework moved earlier | your own estimated cost of the second feed, transfer scheme, or protection change you would have added after the first bad event, counted at the price of building it now rather than retrofitting it later |
| Deal cycle time | negotiation weeks removed x your own carrying cost per week of an open large load negotiation, plus the revenue start date moved forward x your expected monthly revenue from the agreement |
| Emergency restoration | your own cost per emergency callout at a critical customer x the events you judge the design change prevents |
You pay for the modeling service, for the effort to represent the customer's facility, backup generation, and transfer scheme accurately, which requires their cooperation and is the usual sticking point, and for your own engineering and legal time to disposition findings into agreement language. Expect the customer data exchange to take longer than the modeling.
Payback is normally carried by engineering hours and by the design change you make before construction rather than after the first event. The avoided event cost is the biggest line and the least certain, so let it sit as upside.
What is this, exactly? It is AI software: intelligent agents and models built and delivered by SoftServe, running on NVIDIA accelerated computing. It is not a hardware appliance and it does not replace the systems you run today. It deploys in your own cloud or on your premises, connects read-only to your existing systems, and recommends; your people approve every action, starting in shadow mode until it earns trust.
A feasibility and deal-modeling service for the strategy team: for a specific data center inquiry it produces the transmission access analysis, security zone constraint screening, and a draft power purchase agreement (PPA) term structure in weeks. The demo above uses synthetic data; everything below describes what the real deployment needs from your organization.
| Your system | Typical products | How we connect |
|---|---|---|
| Planning and study tools | PSS/E, PowerWorld, TARA | scheduled file export (CSV or CIM XML) |
| Market and grid operator interfaces | PJM, MISO portals; OASIS; settlements | read-only API |
| Geographic Information System (GIS) | Esri ArcGIS Utility Network | scheduled file export (CSV or CIM XML) |
| Document and knowledge stores | tariff books, prior contracts, regulatory filings, SharePoint | document upload |
| Customer Information System (CIS) / billing | Oracle CC&B, SAP IS-U | database replica refreshed nightly |
| Weather and environment | National Weather Service feeds, satellite and LiDAR imagery | read-only API |
| NRC ADAMS, the Nuclear Regulatory Commission's public document database | candidate design licensing documents | read-only API |
Runs in your own cloud account or on an on-premises NVIDIA server; deal terms and grid models are commercially sensitive and access is need-to-know. Security zone screening uses public NRC criteria only, with Safeguards Information excluded by design; outputs are analyses and draft terms, and nothing is committed without executives and counsel.
The Approve button you just clicked in the demo above is the real workflow. This is what it looks like on the screen of the strategy team lead on the hyperscaler inquiry in the GridCORTEX console:
Approve files the feasibility package and draft PPA term sheet to the strategy deal workspace in your document system, as drafts. Legal, licensing, and commercial staff review before any counterparty exchange; GridCORTEX sends nothing outside the utility.
Start by declaring the inquiry: enter the counterparty, requested MW, and target site. The analysis then builds from connected planning, GIS, and market data automatically.
Transmission and market data refresh weekly, GIS and security zone layers on publication; the package cover shows the as-of date of each input.
The deal model lives in the GridCORTEX console; milestone completions notify the team by email and Teams. The console runs in a browser beside your existing screens on day one; embedding into your own systems is a roadmap step once the read-only phase has earned trust. Approve, Modify, and Decline are all captured in an audit trail your compliance team can pull, and GridCORTEX never blocks or overrides anything in the systems you run today.
The fair question from any chief development officer: "We have transmission planners, a land department, outside counsel, and consultants on retainer. What's new here?" Here's the honest answer.
When someone asks "what did it actually calculate?", this is the list. In the simulation these factors drive the storyline; in a pilot they are computed from your grid model, plant records, and interconnection position.
Presenter's one-liner: "An 800 megawatt inquiry landed on a Tuesday, and the old answer was a year-long study the deal would never wait for. The twin was live in a week; the security setback ruled out the north parcel in week two, before anyone spent a dollar on it; three configurations were scored on the same physics in week four, and the hybrid won: a 200 megawatt behind-the-meter bridge plus front-of-meter buildout to 800. Week six: a 20 year term sheet with the refueling outages already priced in. The counterparty's other three sites were still writing consultant scopes."