Off the Books

"In God we trust; all others must bring data."

— W. Edwards Deming

This article is for two readers who do not usually read the same article.

The first is the institutional building owner — the trustee, the facilities VP, the capital projects officer, the school district CFO whose name is on the line for a $10 million geothermal field that has to last fifty years. The second is the infrastructure investor — the GP at a fund that deploys long-duration capital into energy assets and has been waiting, sometimes for years, for the geothermal sector to become bankable.

These two readers are looking at the same asset from opposite sides of a paperwork wall that does not yet exist on either side. The wall is the lifecycle record system. This article is about why the wall has not been built, and what happens to both of you until it is.

Walt's Morning Round

Walt has been a senior maintenance technician at the university physical plant for nineteen years. He starts his Tuesday at 6:42 AM at the central plant, signs into his Maximo dashboard on the tablet, and picks up the day's work orders.

First stop: Engineering Building 3.

He drives across the West Quad parking lot. Three hundred and twelve cars park there on a typical day. Underneath them is the ninety-six-bore field installed in 2018 for the new science building's ground-source heat pump system. Walt does not think about the field. He does not have to. There is nothing in his Maximo for it.

He parks at the loading dock and swipes in. He walks past the chiller plant — six tagged assets each, each chiller, each pump, each VFD, each plate-and-frame heat exchanger. He walks past the switchgear room, every breaker labeled, every panel logged, the last thermographic survey filed in May. He walks past the fire pump, NFPA 25 inspection sticker on the bonnet, last test on file. He passes nineteen smoke detectors on his way to the basement. Each one has a tag.

He pulls work order #87412 — replace pre-filter, AHU-3, MERV-13, 24×24×4. The pre-filter is forty-two dollars retail. He pulls the dirty one, drops it in the recycle bin, slides in the new one, scans the barcode on the new filter into Maximo, and closes the work order.

Twenty-three minutes after arriving, he is back in the golf cart.

By 4 PM he will have passed roughly eight thousand tagged assets across nine buildings. Each one is in a system somewhere. Each one has a record. Each one is, in the language of his industry, managed.

The borefield he drove over twice has none of it.

What "Managed" Means

Every asset Walt touches today is in a system. The institutional building stack underneath any modern campus is some combination of an Enterprise Asset Management platform wired into a Building Automation System and overlaid with an Integrated Workplace Management System. The construction handover into that stack is governed by COBie data drops, ISO 19650 information management protocols, and the asset-management framework codified in ISO 55000 / 55001 / 55002. Facility management itself runs under ISO 41001. Energy management overlays through ISO 50001, which is the document feeding the LEED energy performance prerequisite.

Every asset Walt touches today moves through some part of that documentation chain.

The MERV-13 air filter has a model number, a manufacturer, a unit cost, an installation date, a replacement schedule, a barcode, and a retired-from-service work order that closes the loop on the dirty one. The smoke detector has a serial, a manufacture date, an installation location, a last-tested date, and a replacement-by date. The chiller has a digital twin in the BAS, a maintenance contract logged in Maximo, a refrigerant inventory in the EHS system, and a warranty record in the procurement file. The elevator has a logbook governed by ASME A17.1 and a per-component history that survives the original installer's bankruptcy. The fire extinguisher — fifty dollars retail — has a hydrostatic test record going back to its original certification.

The institutional building owner will track a thirty-dollar smoke detector by serial number, location, last test date, next test date, and replacement schedule for the life of the building.

The same owner will own a five- to fifteen-million-dollar borefield with no asset tags, no per-bore lifecycle record, no servicing schedule, no replaceable-component inventory, and no spatial map that survives the design-build firm's CAD software rotation.

The smoke detector will outlive the borefield's lifecycle record by twenty years.

The Borefield Is Not a Maintenance Asset

The implication of that asymmetry, considered first as a maintenance problem, has a particular shape that the institutional buyer should hold close.

Walt's work order list does not contain the borefield. It cannot. The borefield has no maintenance schedule, because it does not require maintenance the way an air filter or a belt does. It requires maintenance only when maintenance has become an emergency.

The borefield reveals itself through one of two failure modes. Both surface long after the failure began.

The first is a pinhole leak in a loop. For years it releases working fluid into the formation and the void above it. The first sign at the surface is a five-to-ten-year-old depression in the parking lot above the field. By the time the depression matures into a sinkhole, the leak has been undermining the subgrade since the asset was new.

The second is a bad grout job that never sealed the annulus the way the design assumed. Gas accumulated in the formation migrates through the un-sealed annulus, finds the lowest point of escape, and presents at the surface — most often, in an institutional building, by accumulating in a catch basin or sump where the drainage system meets the borefield's footprint. The first sign at the surface is the methane reading.

Neither failure mode appears on a schedule. Both are detected only by their consequence.

This is what a borefield is, considered as a maintenance asset: a system whose condition cannot be inspected, whose components cannot be accessed, and whose failures cannot be predicted. Walt has no way to know, on a Tuesday morning in 2045, which of the ninety-six bores under the West Quad parking lot is leaking.

The most modern installations include RFID tagging on each bore head. This is a useful step. The RFID tells Walt where the bore is. It does not tell him anything else.

It does not tell him what is in the bore — the weights tied to the loop to overcome HDPE buoyancy during installation, the centralizers, the hose fragments, the broken bit teeth, the abandoned tooling. The jewelry of construction, in driller's slang. None of it is documented.

It does not tell him the materials heats, lots, or batches of the pipe, fittings, or grout. A pipe extruded on the same shift as a thousand other pipes shipped to a hundred other sites carries the same potential latent defect. Aerospace manufacturers track this by lot number. Automotive manufacturers track this by VIN. Pharmaceutical manufacturers track this by batch. The borefield's pipe — and the grout that surrounds it — is tracked by nobody.

It does not tell him the construction method on the day his particular bore was drilled. Tremie depth, pump pressure, return-fluid behavior, weather, rig downtime, mud weight at placement — every one of these is a potential correlated failure mode. None of them is recorded against the bore.

The smoke detector Walt walks past on his way to the basement is recallable. If the manufacturer of the model installed in 2015 issues a Class II recall in 2025, every detector with the affected serial range can be identified, ordered out of service, and replaced.

The borefield cannot be recalled. The borefield is not in the system that recalls work.

The downstream consequence the institutional buyer reads immediately: the borefield cannot be warranted. A warranty is a manufacturer's promise indexed to specific units, specific lots, specific construction methods, and specific installation records. The borefield's pipe, grout, and components are not tracked at any of those levels. There is no defect-traceability spine for a warranty to attach to.

Even when a contractor offers a borefield warranty — and the better contractors do — the warranty is voidable on inspection. Most institutional warranties require proof of compliance with certified methods, performed by certified installers, on documented dates, with documented materials. The borefield meets the certified-installer bar in some jurisdictions. It almost never meets the certified-method-and-documented-materials bar. Any warranty claim becomes a paperwork dispute the building owner is structurally positioned to lose.

Aerospace tracks by lot. Automotive tracks by VIN. Pharma tracks by batch. The borefield is tracked by nobody.

What That Costs, Past the Maintenance Bay

The maintenance gap is the gap the institutional buyer feels first. The deeper costs surface in three places past the maintenance bay.

Insurance. Carriers price what they can underwrite. Insurance carriers writing coverage for institutional energy assets — equipment breakdown, contamination liability, performance guarantees — increasingly require documented asset histories as a condition of coverage. A borefield with no lifecycle record is uninsurable except at premiums that wipe out the asset's economics. The carrier's logic is not malicious. It is the same logic the warranty regime uses: if the underwriter cannot trace the asset to its construction, the underwriter cannot price the risk. An untraceable asset is treated, for underwriting purposes, as worst-case.

Sale and disposition. Public schools, hospitals, universities, and government buildings are routinely sold, demolished, or repurposed. The asset that should fetch the highest price — a working borefield with a fifty-year design life and decades of remaining capacity — fetches nothing if it cannot be transferred with documented integrity. Buyers cannot underwrite an unknown. Public-asset disposition processes that depend on appraised value will appraise the borefield as a liability, not an asset, when the records do not exist. The institution that paid ten million dollars for the field cannot recover anything when the building's mission changes. The next owner buys the building and inherits a fifty-year question they cannot answer.

Infrastructure capital — the structural blocker. This is the largest stake in the article, and it deserves to be named directly.

The energy transition needs trillions of dollars of infrastructure financing. Geothermal — particularly closed-loop ground-source heat pumps at institutional scale — is one of the largest unaddressed opportunities in the building sector. Infrastructure investors at the major funds (Macquarie, Brookfield, Blackstone Infrastructure, KKR Infrastructure, Stonepeak, the public-sector pension funds, the insurance company general accounts) deploy long-duration capital into assets they can underwrite, transfer, and refinance.

Their underwriting requires documented asset records spanning the lifecycle, performance guarantees backed by data the asset itself produces, insurable risk profiles, and transferable ownership. A borefield without lifecycle documentation fails every one of those tests.

The capital that would otherwise be eager to fund the GSHP buildout — at scale — cannot deploy, because the asset is not bankable. Geothermal as a Service models, third-party-owned thermal utilities, energy-as-a-service contracts, green bonds tied to thermal infrastructure — all require the documentation layer to exist before the financial layer can be built on top.

The reason geothermal has not scaled the way solar PV has, despite the engineering being better understood in many ways, is that solar PV is documented, metered, and transferable from the first day of operation. Every panel has a serial number. Every inverter reports to a monitoring platform. Every PPA has a measurable performance baseline. The borefield has none of those things.

The structural blocker on geothermal at scale is documentation. The asset class does not yet exist as an asset class because the records that would constitute the asset class do not yet exist.

What the Standards Already Say

ISO 55000 / 55001 / 55002 codifies what asset management looks like. ISO 41001 codifies what facility management looks like. ISO 19650 codifies what BIM-based lifecycle handover looks like. COBie codifies what data the contractor must hand to the owner at substantial completion. The four standards together describe a closed loop: the construction record becomes the operating record becomes the disposition record becomes the underwriting record.

For every other major component in the building above the borefield, that loop is closed by default. For the borefield, the loop has never been closed because the construction record does not generate the data the operating record needs.

C449, the bore-construction standard now being drafted, is the document that closes the loop on the construction-record side. It defines what data a contractor must generate for each bore, in what format, with what verification, and in what handover package. Once C449 is adopted by reference into building codes, LEED submissions, procurement specifications, and insurance underwriting, the construction record will exist for the borefield the same way it exists for the chiller plant. The asset becomes warrantable. The carrier prices the risk. The pension fund underwrites the deal. The capital that has been waiting on the asset class arrives.

What That Means for Both Readers

Walt drives past eight thousand tagged assets on a Tuesday. The single largest asset on his campus has none. Until that asymmetry is closed, the borefield is a maintenance burden in the present, an insurance liability in the medium term, a write-off at disposition, and an asset class that capital markets will not enter at the scale the energy transition requires.

The institutional buyer asks: what do I do with the borefield I already own? The answer is to retrofit the lifecycle record, today, against the bores that are already in the ground. The records can still be reconstructed for many of them — for some, only partially, but partial is closer to bankable than nothing. The cost of reconstruction is dwarfed by the cost of the borefield being treated as a write-off in five, ten, or twenty years.

The infrastructure investor asks: when does this asset class become deployable? The answer is when the construction record becomes a standard handover, the operating record becomes a continuous data stream, and the disposition record becomes a transferable asset history. C449 closes the construction-record gap. The platform that turns each bore into a digital twin closes the operating-record gap. The two together turn the borefield into something the carrier can underwrite and the fund can finance.

Walt walks past eight thousand assets that can be warranted. He drives over the one that cannot.

The fix is the standard, the platform, and the discipline of treating the bore the way every other asset in the project is already treated.