It is 9 a.m. on commissioning day. The ground modules are in the trench, the grid is connected, and the consultant is standing over your shoulder with the acceptance sheet open to one line: measured earth resistance. The datasheet that won the tender says the modules deliver a low, season-stable resistance. The instrument in your hands is about to decide whether that was a specification or a slogan. After twelve years on the supplier side of this moment, I can tell you the test goes wrong for predictable reasons — wrong spike spacing, a reading taken in the wrong season’s assumptions, or a method chosen for convenience instead of validity — and almost never because the physics surprised anyone. This guide walks QA engineers and commissioning teams through the field methods that actually verify a ground module’s resistance, how to place the test spikes, how to judge whether a reading is trustworthy, and what to demand from your supplier so that the promised number and the measured number are the same number. Our own modules ship with the parameters you need for that verification, and I will show you exactly where they plug in.

Why a Datasheet Resistance Is a Promise, Not a Measurement

Every ground module datasheet — ours included — states resistance performance measured under defined conditions: a stated soil resistivity, a stated installation geometry, a stated test method. That number is honest only inside those conditions, because earth resistance is not a property of the module alone; it is a property of the module plus the soil it sits in plus the way it was buried. Move the same module from the factory test plot to your rocky, dry site and the number moves with it. I have watched the same module pass in our test conditions and read multiples of that value on a client’s dry ridge — same product, same lot, different soil, and that gap is exactly what field verification exists to catch. This is not a defect in the datasheet — it is what the datasheet is for. The supplier’s job is to state the per-module resistance and its basis so your engineer can model your site; the commissioning team’s job is to verify the result in your soil, with your installation, on your acceptance date.

I tell buyers to treat the supplier’s resistance figure as the input to a calculation, never as the output. When our clients model a site, we provide the per-module resistance for the selected form — plum-blossom, column or square — together with the soil conditions behind it, because a figure without its basis cannot be scaled to another site honestly. The verification that closes the loop happens in the field, and the rest of this guide is about doing that verification in a way that stands up to an auditor, a consultant, and a second visit a year later.

Fall-of-Potential Done Right: Where Do the Test Spikes Actually Go?

The fall-of-potential method remains the reference measurement for earth resistance, and it fails more often from geometry than from instruments. The logic: inject a test current between the electrode under test and a remote current spike, then walk a potential spike along the line between them and read the voltage profile. The flat part of that profile — where the reading stops changing — is the electrode’s true resistance zone, and the classic shortcut places the potential spike at 61.8% of the distance to the current spike, with readings at 52% and 72% as confirmation. When I train a client’s crew on this, we run the full profile once before trusting the shortcut, Because a flat-zone confirmation is what separates a measurement from a ritual. The full procedure and its assumptions are the subject of IEEE’s grounding measurement guidance built around IEEE Std 80, and the underlying soil model comes from a Wenner-style resistivity survey per ASTM G57.

The spacing discipline is where field tests quietly go wrong. If the current spike sits too close, the electrode’s and the spike’s resistance zones overlap, the voltage profile never flattens, and the 62% reading comes back falsely low — the test passes a module that the soil has not. I tell commissioning teams the rule of thumb our engineers use: the current spike distance should be at least five times the largest dimension of the electrode installation, and more in high-resistivity ground, because the dry, resistive soil that justified buying modules in the first place also stretches every resistance zone in the test. Crowd the spikes and you verify nothing.

A column carbon ground module of the type verified by fall-of-potential testing at handover — the terminal stud connects to the test instrument via the grid lead.

Three Field Methods, One Acceptance Sheet: Which Test Fits Your Site?

Fall-of-potential is the reference, but it is not always available — live sites, cramped compounds and connected grids force alternatives. The honest comparison below is the one I send to QA teams who ask which method their consultant will accept. We keep all three method briefs on file, and I share them with the client’s consultant before commissioning week so nobody debates method validity beside an open trench. The column that matters most is the last one: what each method actually proves.

Two cautions from our side of the fence. First, the clamp-on meter is the most misused instrument in earthing: it is excellent at confirming loop integrity on a bonded system and physically incapable of testing a stand-alone module, yet we see commissioning reports every year where it was pointed at an isolated electrode and produced a confident, meaningless number. Second, whichever method you use, record the geometry — spike distances, soil condition, weather — next to the reading. Because a resistance value without its test geometry cannot be reproduced, an unannotated pass is a pass you cannot defend when the client re-tests at the first anniversary.

Reading the Number: When Is a Test Result Trustworthy?

A trustworthy reading has four attributes, and I teach our clients’ QA teams to check them in order. Geometry: the spike spacing met the validity rules above. Season: the reading was interpreted against soil moisture at test time — a module tested in the wet season should show margin against the target, because the dry season will reclaim some of it unless the module’s moisture-holding design is doing its job. Stability: repeat the measurement after disturbing and re-seating the potential spike; a stable profile means the flat zone is real. And documentation: instrument model, calibration date, geometry and weather recorded with the value. A reading with all four is evidence; a reading missing any one is a number. In our own export projects I ask the commissioning team to send me the annotated sheet, and we file it against the production lot on our side.

For graphite ground modules specifically, the verification logic has one pleasant asymmetry worth exploiting — and I point it out to every QA team we work with. The module’s conductivity is season-independent by construction — graphite conducts as a solid — and its moisture-absorption behavior holds the soil interface tight, so the wet-to-dry drift that haunts bare-rod readings is exactly what the product suppresses. If your wet-season and dry-season readings on a module point come back nearly identical, that is the design working, not the instrument failing. Our clients in arid regions tell me the second-season re-test is where the product earns its reorder, and I believe them — we see the reorders.

Square-form module bedded for testing: shallow-pit installations like this are verified with the same fall-of-potential geometry, with D sized to the array dimension.

Supplier Verification: What Should You Demand Before the Modules Leave the Factory?

Field testing catches problems late. The sharper tool is supplier verification before shipment, and here is the five-point demand list I recommend to QA engineers — it is also, deliberately, the list our own export orders answer:

1. Per-module resistance with its test basis — the value, the soil resistivity behind it, the installation geometry and the method, stated together or not at all.
2. Lot-traceable production records — our modules ship from a 10,000-piece monthly line under ISO 9001 discipline with incoming (IQC) and process inspection; the certificate should tie to your lot, not to the catalog.
3. Impulse-current behavior in writing — our production characteristics state no resistance increase, no hardening and no brittle fracture after large impulse current shock; ask any supplier to put the equivalent claim in the contract documents.
4. Material declaration — nonmetallic graphite body, corrosion-resistant, non-toxic, non-polluting; for projects with an environmental reviewer this declaration travels with the submittal, consistent with the earth-electrode component scope of IEC 62561-2.
5. Acceptance criterion agreed before shipment — target value, test method and spike geometry written into the order, so commissioning day holds no negotiation about what “passed” means.

Point five is the one that separates verification from theater. When the acceptance criterion is agreed at order time, commissioning becomes a measurement; when it is not, commissioning becomes an argument. Because we quote against stated per-module resistance and stated test geometry, our clients arrive at the acceptance test with the calculation already done — the instrument only has to confirm it. Designs that follow IEC 62305-1 for lightning protection should carry the same discipline into the LPS earth-termination records.

From Test Report to Acceptance: How Do You Close the Loop?

The closing step is administrative, and it protects everyone — I have seen it rescue a claim and I have seen its absence sink one. File the measured values with their geometry and conditions, reconcile them against the pre-agreed acceptance criterion, and archive the set with GPS tags per point so the first-anniversary re-test measures the same points the same way. I keep this archive habit in our own shipment documentation, and I recommend it to every QA lead we support. When a reading fails, the diagnosis sequence is: verify spike geometry first, verify instrument and leads second, excavate and inspect bedding contact third — in that order, because test setup errors outnumber product failures in the field reports I have reviewed by a wide margin. And when a supplier’s module genuinely misses its stated per-module resistance under the agreed method, a lot-traceable record is what makes the claim resolvable in days instead of a dispute that outlives the project.

Terminal detail on our graphite module: a clean instrument connection point matters more than most teams expect — corroded or loose test leads are a classic source of phantom high readings.

The products behind this guide: our ground module family (AF-G0701) in plum-blossom, column and square forms, the column graphite ground module for bored-pocket installations, and the carbon ground module with low resistance (AF-G0705) — all under the ground module category, from ten-piece minimum orders with wooden-case export packing. For the material behavior behind the test numbers, our companion article on graphite electrodes and their role in grounding covers why a graphite body holds its conductivity where metallic and fill-based solutions drift.