7/23/2019 Boletin 62.8%

1/2

NETA WORLD Spring 2006 1

In previous columns, we have examined the familiar and popular

methods for testing the resistance of electrical grounding electrodes.

To the uninstructed, these methods may at rst appear overly detailed,

tedious, and time-consuming. They typically involve careful measuring

and positioning of test probes, repeated test measurements, and often

graphing and/or mathematical proofs. Compared to operating a multi-

meter, this can look like too much work!

But test methods exist and are precisely described, because it is not theresistance of the electrode itself that is being measured, but its electricalrelationship to the surrounding earth. The test item isnt so much theelectrode as it is the entire planet Earth or, ...for practical applications...

, that portion immediately surrounding the buried electrode. This is aradical divergence from testing a piece of equipment, where circuitry iswell-dened and current ow reasonably predictable. In routine electricaltesting, the test meter is connected so as to become an extension of thecircuitry, and the test is fairly comprehensible. A voltmeter, for instance,

bridges the load and measures the electro-motive force across it. In mostelectrical equipment, the better the current can be controlled and directed,the more efcient the device becomes. But grounding is the opposite.Since the electrodes function is to divertcurrent (from the facility), theless restricted it becomes, the better. The earth fullls this requirementadequately by its enormous volume, but that also makes measuring it achallenge. Hence, the need for specic and well-dened procedures.

Test methods exist because there are so many challenges to be ad-dressed such as poor soil, huge electrodes with large electrical elds, andlimited and highly obstructed workspaces. Over many years of eldwork,different methods have been devised to meet these varying challenges.Methods serve the purpose of proving the result and/or saving timeand reducing workload. Some are also adapted to dealing with specicproblems, particularly limited space. Some methods, like fall-of-poten-tial, provide excellent proof of reliability but at the expense of time andwork. Others, like the 62 percent method, save time, but the result must

be accepted without proof. Still others, like simplied fall- of-potential,compromise by reducing test time, yet still provide a (mathematical) proofof reliability. Methods designed to work in limited space, because of the

Ground Resistance Testing:Four Potential Method

inherent danger of getting a falsereading with restricted lengths of

test leads, must by their nature in-clude some means of proof, wheth-er mathematical (slope method) orgraphical (intersecting curves).

All of these methods have beendescribed in detail in earlier col-umns. In this issue, we will exam-ine the four potential method. Thisis another method specically in-tended to meet the problem of test-ing large ground grids that wouldotherwise require impracticallylong test leads and, therefore, is fre-

quently employed for substationsand power stations. Such gridstend to cover large areas and alsohave complex structures of differ-ent sized rods. When approached

by conventional methods, not onlymay there be insufcient space toaccommodate the required lengthof test leads, but the electrical cen-ter of the system is indeterminate.Trying to use a method like the 62

Tech Tips

by Jeff JowettMegger

7/23/2019 Boletin 62.8%

2/2

Article available for reprint with permission from NETA. Please inquire at neta@netaworld.org

2 NETA WORLD Spring 2006

percent rule, which is based on fairly precise measure-ment for positioning of the test probes, would intro-duce too large a potential error. The leads would haveto be measured from an arbitrary point, which could

be a considerable distance from the electrical center.With four potential, the tester is attached as usual toany convenient point on the edge of the groundingsystem. The current probe is stretched out to what-ever maximum convenient working distance can be

obtained, but it is still advisable to do all that is pos-sible in order to gain a substantial span.

Figure 1 Four Potential Method Test Confguration.

Critical readings are then taken with the potentialprobe at 0.2, 0.4, 0.5, 0.6, 0.7 and 0.8 the distance to thecurrent probe (Fig. 1). If the local situation followedthe ideal model, these points could simply be plottedinto a fall-of-potential graph, and the job would bedone. But four potential works where fallof-potential

would produce an ambiguous and unreadable graph.A mathematical proof accomplishes what a graphiccannot. It determines at what point the resistanceof the test electrodemaximizes while masked by thesuperimposed resistance associated with the current

probe. This could also be visualized as the theoretical62 percent point if the test ground were offering anideal model. The resistance values determined at thesix positions (designated R

1through R

6) are linked

through four formulae to the true resistanceat in-nite distance (i.e., if it were possible to measure theplanet as a whole), as follows:

(a)R= -0.1187 R

1 0.4667 R

2+ 1.9816 R

4 0.3961 R

6

(b)R= -2.6108 R

2+ 4.0508 R

3 0.1626 R

4 0.2774 R

6

(c)R

= -1.8871 R2+ 1.1148 R

3+ 3.6837 R

4 1.9114 R

5

(d)R

= -6.5225 R3+ 13.6816 R

4 6.8803 R

5+ 0.7210 R

6

It is expected (and hoped!) that these four calculatedresistances will essentially agree and can then be av-eraged to produce the nal test result. Because of theunderlying assumptions in the theory upon which themethod is based it may happen that the calculationresulting from equation (a) may notbe in agreementIf so, it can simply be disregarded and the averagetaken on the remaining three. Because of the numberof variables involved and the enormous range of pos-

sible soil resistivities, it would be a prohibitive task totry to scientically dene essential agreement. Forpractical eldwork, the operators work experienceis always an essential ingredient. Familiarity with theprocedure can be counted on to develop a sense owhen its working and when its not.

There still may remain a complication based ondistance to the current probe. In order for all of theunderlying assumptions to be valid, the theory behindfour potential still requires a fairly long lead lengthto the current probe. For typical larger substations, orthose with exceptionally low ground resistances, theselengths can run to 1000 or 2000 feet! This may not bepossible, yet if not achieved, the resultant equationswill fail to produce satisfactory agreement. All has no

been lost, however! Failure of agreement at least informs the operator that a valid test has notbeen madeSome time has been expended, with the mathematicaltest proong the result, whether acceptable or notIn the latter case, the test can be repeated using an alternate method that can be successfully accomplishedwith shorter lead lengths. Favorites in such instancesare the slope and intersecting curves methods, whichhave been described in earlier columns.

The theory upon which the four potential method is

based is complicated, relying heavily on mathematicaderivation and assumptions about the shape of the teselectrodes electrical eld. It will be explored in moredetail in a future column.

ReferencePaper 4619S from the proceedings of the IEE vol. III

No. 12, Dec. 1964 by Dr. G. F. Tagg.

Jeffrey R. Jowett is Senior Applications Engineer for Meggein Valley Forge, PA, serving the manufacturing lines of BiddleMegger, and Multi-Amp for electrical test and measuremeninstrumentation. He holds a BS in Biology and Chemistry from Ursinus College. He was employed for 22 years with James G. BiddleCo. which became Biddle Instruments and is now Megger.