Earthing resistance for data centers: the 5 ohm rule and when sites ask for less
PUIL is widely cited for a maximum earth resistance of 5 ohms, and that figure is the usual starting point on Indonesian projects, but many data center specifications ask for a lower value and the project specification is the one you test against. Earth resistance is measured with the fall-of-potential method, and a reading above the target is brought down by adding or deepening electrodes, extending the grid or treating the soil, then testing again.
What earth resistance is and why a data center cares
Earth resistance is the resistance between the earthing system, whether a single rod, a group of rods or a buried grid, and the mass of ground around it. When a fault puts current onto a metal enclosure, that current needs a low resistance path so the protective device sees enough of it to operate quickly, and so the voltage on the enclosure while the fault lasts stays low enough that anyone touching it is safe.
A data center leans on its earthing system for more than fault clearance, since the lightning protection, the surge protective devices and the bonding network under the raised floor all discharge through it, and a poor earth lets voltage differences build up between equipment that is meant to sit at the same potential.
The 5 ohm figure and how it is used
PUIL is widely cited for a maximum earth resistance of 5 ohms, and many Indonesian drawings and site procedures use it as the default target. Many data center specifications ask for a lower value, sometimes for the whole earthing system and sometimes for one part of it such as the lightning protection or the transformer earths, and where they do, the specification governs.
Read the earthing clauses of the specification before the electrodes go in rather than after, because the target decides how many electrodes the design needs, and adding them while the excavation is open is far easier than breaking into a finished slab.

How earth resistance is measured
The usual method is fall-of-potential, also called the three-point method, in which the tester passes a current between the electrode under test and a current probe driven into the ground some distance away, then measures the voltage between the electrode and a potential probe placed between the two, and the ratio of voltage to current gives the resistance. Readings are taken with the potential probe at several positions, and the result is only trusted when the readings stay steady as the probe moves, which shows the probes are far enough apart.
Where possible the electrode is disconnected from the installation for the test, because otherwise the current also flows through cable sheaths, the incoming supply earth and every other bonded path, and the reading describes the whole network rather than the electrode. We record the soil conditions, the recent weather and the probe positions with each result, so a later test can be compared on the same basis.
Why readings vary
Earth resistance depends mostly on the resistivity of the soil around the electrode, which changes with soil type, moisture and temperature, so clay and wet ground read lower than sand, gravel or rock, and an electrode that reaches a permanently moist layer stays steadier through the year than a shallow one. Seasonal moisture is the largest variable on most Indonesian sites, which is why the test date and soil condition belong on the test sheet beside each reading.
The measurement itself can also mislead, since probes set too close to the electrode or near buried pipes, cables or reinforcement give a figure that does not represent the electrode, and repeating the test with the probes laid out in another direction is the simplest check.
What to do when the reading is too high
First confirm the reading by checking that the electrode was isolated and the probes were well placed, then repeating the test. If the value is still above the target, the usual remedies are to add electrodes, drive them deeper to reach wetter or more conductive soil, extend the grid or join separate electrode groups together, or treat the soil around the electrodes with a conductive backfill to lower its resistivity.
Each remedy changes the earthing system, so the designer should approve it before it is built, and the electrode is tested again afterwards with the new result recorded beside the first one. Soil treatment can lose effect over time, so where it is used, ask whether the specification calls for periodic retesting.
Questions about earthing resistance
Is 5 ohms the maximum?
It is the figure widely cited from PUIL and the usual default, but it is not the target on every project, because many data center specifications ask for less. Treat 5 ohms as the upper limit only when the specification sets no value of its own.
Do IT loads need a lower value?
The specification decides, and data center specifications often set a lower value for the whole system or for parts of it, such as the earth serving the data hall. There is no single figure for IT loads, so test each part against the value the specification names for it.
How is earth resistance measured?
With an earth resistance tester and two auxiliary probes, using the fall-of-potential method. A clamp-on tester is sometimes used on electrodes that form part of a larger earthing network, but it reads the whole loop rather than the electrode alone, so the specification should accept that method before it is used.
Is earth continuity a separate test?
Yes, continuity checks that every exposed metal part, from switchboards to cable trays and raised floor pedestals, is bonded to the earthing system, while earth resistance measures the electrode against the ground. A site needs both, since a good electrode does nothing for a cabinet whose bond is missing.
Why does the reading change between dry and wet season?
Current flows through the ground mainly by way of the water and dissolved salts in the soil, so as the soil dries its resistivity rises and the electrode reads higher. A result taken in the wet season can therefore look better than the system will perform in a long dry spell.
Book electrical testing
Send the single-line diagram, the specification and your target dates, and we'll reply with a test plan and a quote.
