Power and resource

You cannot sell a kilowatt hour you cannot measure.

Constanellis is a power provider. The same fact that makes surface power hard to build makes it impossible to sell honestly: there is no instrument on any lunar surface that can say how much energy crossed an interface, so there is nothing to bill against and nothing to hold a supplier to. We are building the provision and the instrument together, because neither is worth much without the other.

Nothing on this page is sold today. Every line carries the stage it is actually at, and the resource work carries no line at all because none is offered.

The problem underneath

Two sides of a power interface routinely state different numbers for the same connection.

One side plans to deliver. The other expects to draw. On the ground the difference is absorbed by margin, and nobody looks. On a surface where every watt arrived at launch cost, the difference is the whole commercial question, and the first time anyone looks at it is the first time it fails.

A headline efficiency figure has the same defect. It is only useful when the system boundary, the operating conditions, and the treatment of uncertainty are stated with it, and they usually are not. A number without those three is not a measurement, it is a claim in the shape of a measurement.

It is not only the numbers. There is no conformance scheme for a lunar power interface either. The governing interface document states that the physical interfaces between loads and power systems are not defined, and describes its own procedures as suggested. So two parties can both comply with it and still fail to connect, and there is no test anyone can point at to find that out before the hardware is on a surface.

So a power business on the Moon has a sequencing problem that nobody has solved. You cannot write a supply agreement against delivered energy until delivered energy is measurable. You cannot arbitrate a dispute between a provider and a consumer without a party who measured and has no stake in the result. And you cannot price energy at all until there is a verified quantity to price.

The instrument is the product. The power is what the instrument makes sellable.

6 lines

What we provide, and the stage each line is at.

These are the WIRES lines exactly as the master catalogue carries them. Nothing here is sold today, and the prerequisite column says what has to exist before each line can start.

Line	What you receive	Unit of sale	Status	Prerequisite	Earliest
W-01Power management and distribution units	Conversion, regulation, protection and switching hardware qualified to the interface it serves	Per unit	In development	Design and qualification	Year 2
W-02Surface power distribution harness	Cable, connector and protection design for distance runs at voltage, with the loss budget measured rather than calculated	Per run, priced by length and rating	In development	Design and qualification	Year 2
W-03Power interface verification and acceptance measurement	Independent measurement of what actually crossed the interface, against what each side expected to deliver and receive	Per interface, per campaign	In development	Instrumentation	Year 1 to 2
W-04Independent measurement referee, power beaming and radio frequency	A measurement by a party with no stake in the result, which is what an acceptance test is for	Per test campaign	In architecture	Instrumentation and method	Year 2 to 3
W-05Power provision against a delivered energy agreement	Energy delivered and verified, billed on the verified quantity	Per kilowatt hour delivered and verified	In architecture	An instrument that does not yet exist anywhere	Gated
W-06Energy storage characterisation	Capacity, rate and degradation measured across the temperature range the mission actually sees	Per cell, module or string	In development	Facility	Year 2

Every line in the catalog, across all programs

How power is sold

A delivered energy agreement, billed on the verified quantity.

The unit of sale is a kilowatt hour delivered and verified, not a unit of hardware and not a nameplate rating. That is a deliberate choice and it is expensive: it puts the measurement risk on the provider, where a customer would want it, and it means the company cannot invoice for energy it cannot prove arrived. The catalogue records the consequence honestly. Line W-05 is in architecture, its prerequisite is an instrument that does not yet exist anywhere, and its earliest start is gated.

What is metered: Energy across a defined interface, under stated operating conditions, with the boundary written before the meter is installed rather than after the invoice is disputed.
What is billed: The verified quantity. Energy that was delivered but cannot be verified is not billed, which is the only version of this agreement a buyer should accept.
Who arbitrates: A measurement by a party with no stake in the result, which is what an acceptance test has always been for. Catalogue line W-04 is that service sold on its own, to parties who are not buying our power.
What is not promised: Availability, continuity, and uptime are not offered against any surface asset, because no surface asset exists. A delivered energy agreement is a commercial structure the company is designing, not a service a customer can buy today.

What nobody publishes

A power transfer figure without these eight fields cannot be compared to anything.

Nothing in the public record fixes how a power transfer result must be reported. The nearest thing is a survey in the peer reviewed literature which found reported results historically inconsistent and proposed a reporting scheme, and no standards body has adopted it since. The practical effect is that a buyer comparing two suppliers is comparing two announcements: different boundaries, different conditions, different denominators, and in several cases no stated distance at all.

So we are publishing the fields first, before we have a customer for the measurement and before we have a result of our own to report. These are the eight things a power transfer claim has to carry to be checkable. We will fill in every one of them for our own results. That is the only reason it is reasonable to ask anyone else for them.

The system boundary

Where the measurement starts and where it stops, named on both sides. A figure quoted end to end and a figure quoted across one subsystem are different figures, and nothing in the number itself says which it is.

The denominator

What the result is a fraction of. Electrical input to electrical output is not optical output to electrical output, and the two are routinely reported under the same word.

Reference conditions

Distance, aperture, alignment, temperature, pressure, and the state of anything tuned for the run. A result taken at a shorter range than the headline range is a different result, and the range belongs with the number rather than in a footnote.

The loss list

Every loss term between the two ends of the boundary, named, with which were measured and which were calculated marked separately. A loss nobody listed is a loss nobody measured.

Uncertainty

Every term with its type, its distribution and its basis, combined and expanded, with the coverage factor stated. A result without an uncertainty is not comparable to a result with one, in either direction.

Duration and duty

How long the run held, and whether the figure is a peak, a mean over a stated window, or a single sample. A peak reported as a capability is the most common way this market overstates itself.

Who measured, and with what

Whether the measuring party had a stake in the result, the instrument chain used, and when each instrument in it was last calibrated against what.

What this does not establish

The claims the run does not support. A result that establishes nothing about scaling, endurance or the environment should say so, and the party who ran it is the party who knows.

The same eight questions rewrite cleanly for a resource process: boundary, denominator, feedstock and conditions, loss and energy list, uncertainty, duration, who measured, and what it does not establish. A yield figure has exactly the comparability problem a transfer figure has, and for exactly the same reasons.

This is a reporting template, not a standard and not an accreditation. We are not a standards body and we do not issue certificates. It is the form of a claim we will answer, and the form of a claim we will publish.

The resource half

A resource is not a resource until somebody has measured it.

Every lunar resource argument runs on numbers that disagree with each other. How much water is in the cold traps, how deep, and in what form: the public estimates differ by more than an order of magnitude, and a business case cannot be written across that spread. Oxygen is roughly forty percent of regolith by mass and locked into oxides, which is a chemistry problem sitting on a thermal problem sitting on a power problem. None of that is settled, and the site will not pretend otherwise.

Our position in resources is the measurement, not the mine. Characterising a resource process is the same discipline that characterises an optic in vacuum: define the boundary, state the conditions, budget the uncertainty, keep the record checkable. That is a service we will sell to people running extraction processes long before we run one, and it is why the resource work sits next to the power work rather than in a separate future.

None of this carries a catalogue line, because none of it is sold. The four threads below are research and architecture, and each one states what is not claimed.

Oxygen from regolith

Studied as part of a materials process rather than as a dedicated plant, on the argument that a process which yields oxygen and a usable solid is worth more than one that yields oxygen alone. The measurement question comes first: there is no accepted benchmark for stating a yield figure, so every published figure is hard to compare.

What does a kilogram of oxygen actually cost in energy, at what rate, from which feedstock, under which thermal conditions?

No extraction process is operating, no yield has been demonstrated by us, and no oxygen figure appears on this site.

Regolith as feedstock

Treated as a qualification problem rather than a printing problem. A part made on the surface is worth nothing until somebody can characterise it, and the characterisation chain is the thing we already build. Regolith manipulation at scale sits on the published technology gap lists as a named challenge.

What has to be true before a part made on the surface can be flown on something that matters?

Nothing has been manufactured from regolith by us. This is a research thread with no product and no date.

Volatiles and the water question

Handled inside the measurement corpus rather than as a mission. Where public estimates disagree by more than an order of magnitude, the useful contribution is a comparison frame that says which results can be placed side by side and which cannot, and the dependency register already carries the disagreement rather than resolving it.

How do you compare two prospecting results that used different instruments, boundaries, and assumptions?

We have not prospected, and we do not hold an independent estimate of lunar water inventory.

Energy across the night

Catalogue line W-06, energy storage characterisation: capacity, rate, and degradation measured across the temperature range the mission actually sees. Surviving the night is the resource question and the power question at the same time, and the published phase-one target of more than 120 hours is a fraction of the real one.

What does a storage system actually do across about 354 hours of darkness, as opposed to what its datasheet says at room temperature?

In development. No storage system has been characterised by us across a lunar night or a lunar night analogue.

Read this before the rest

What this offering does not claim.

No power service is operating. No energy has been delivered, measured, or billed under any of these lines, and none is offered for sale today.
No efficiency, distance, yield, or power figure for any system appears on this site, ours or anyone else's, because none has been measured by us.
No resource extraction process is operating, and no resource line appears in the catalogue because none is sold.
A measurement definition is not an accreditation, and a comparison frame is not a ranking. We sell measurement and evidence. A certificate we are not accredited to issue is not for sale.
Nothing here has flown.

Engagement

How an engagement runs.

The lines in development and in architecture are not ordered. They are scoped, and scoping is free.

Frame
You describe the system at a non-sensitive level, the performance question, and the conditions you need to understand. No technical data changes hands.
Scope
We reply with a proposed boundary, the conditions to state, and the terms an uncertainty budget would have to carry. That reply is useful whether or not you ever buy anything from us.
Agree
Scope, conditions, and the handling of any controlled data are agreed in writing before measurement work is planned. Controlled technical data is handled under our ITAR registration and does not move before that agreement exists.

Bring the interface, not the part number.

The useful first conversation is about a boundary and a disagreement: where the interface sits, what each side expects to deliver and to receive, and which number the two sides do not agree on. That is a scoping conversation, and it starts without an order.

Scope a measurement The WIRES program