to the 16th Edition IEE Regulations
   
   
   
 
 

chapter 5
Earthing

chapter 6
Circuits

Inspection and Testing
  8.1 - Introduction 8.5 - Insulation tests
  8.2 - Inspection 8.6 - Earth testing
  8.3 - Testing sequence 8.7 - Test instrument requirements
8.4 - Continuity tests 8.8 - Supporting paperwork


8.6.2 - Measuring earth-fault loop impedance and
prospective short-circuit current

The nature of the earth-fault loop and its significance have been considered in detail in {5.3}. Since the loop includes the resistance of phase and protective conductors within the installation, the highest values will occur at points furthest from the incoming supply position where these conductors are longest. A measurement within the installation will give the complete earth-fault loop impedance far the point at which it is taken (Zs), or the earth-fault loop impedance external to the installation (Ze) may be measured at the supply position. Internal loop measurements should be taken at points furthest from the intake to give the highest possible results.

In simple terms, the impedance of the phase-to-earth loop is measured by connecting a resistor (typically 10 Ohms) from the phase to the protective conductor as shown in {Fig 8.17}. A fault current, usually something over 20 A, circulates in the fault loop, and the impedance of the loop is calculated within the instrument by dividing supply voltage by the value of this current. The resistance of the added resistor must be subtracted from this calculated value before the result is displayed. An alternative method is to measure the supply voltage both before and whilst the loop current is flowing. The difference is the volt drop in the loop due to the current, and loop impedance is calculated from voltage difference divided by current.

Fig 8.17 - Simple principle of earth-fault loop testing

Since the loop current is very high, its duration must be short and must be limited to two cycles (or four half-cycles) or 40 ms for a 50 Hz supply. The current is usually switched by a thyristor or a triac, the firing time being controlled by an electronic timing circuit It is very important to have already checked the continuity of the protective system before carrying out this test. A break in the protective system, or a high resistance within it, could otherwise result in the whole of the protective system being directly connected to the phase conductor for the duration of the test. Commercial testers are usually fitted with indicator lamps to confirm correct connection or to warn of reversed polarity. {Fig 8.18} shows a typical earth-fault loop tester connected to a socket outlet so that its loop impedance can be measured. If the circuit to be measured includes socket outlets, the tester is connected as indicated in {Fig 8.18}. Special leads for connection to phase and to earth are provided by suppliers for all other circuits.

Fig 8.18 - Earth-fault loop tester connected for use

Before testing, the main equipotential bonding conductors are disconnected (BUT NOT THE CONNECTION WITH EARTH) to prevent parallel earth return paths and to ensure that there is no reliance on the service pipes for gas and water for effective earthing, (REMEMBER TO RECONNECT THE MAIN EQUIPOTENTIAL BONDING AFTER THE TEST).

Tests must be carried out at the origin of the installation, at each distribution board, at all fixed equipment, at all socket outlets, at 10% of all lighting outlets (choosing points farthest from the supply) and at the furthest point of every radial circuit. The test should be repeated at least once to allow for the effect of transient variations in the supply voltage.

A modified version of the earth-fault loop tester, which effectively measures the phase to neutral impedance and calculates then displays the value of the current which would flow if the supply voltage were applied to this impedance are readily available. The principle of such a PSC tester is described in {3.7.2}.

Since the test result is dependent on the supply voltage, small variations will affect the reading. Thus, the test should be repeated several times to ensure consistent results. The test resistor will be connected across the mains for the duration of each test. and will become very hot if frequent tests are made. Some testers will then 'lock out' to prevent further testing until the resistor temperature falls to a safe value.

      The earth fault loop impedance measured as described will be for installation cables at ambient temperature, unless the circuit concerned has been in use immediately before the test, when it will be the impedance at normal operating temperature. Under normal operating conditions, cable temperature will rise, and so will the resistive component of the impedance. This effect is difficult to calculate, and a practical alternative is to ensure that the measured values of earth fault loop impedance do not exceed three quarters of the maximum values shown in {Tables 5.1, 5.2 or 5.4} as appropriate.

The effect of supply voltage on the calculation of earth fault loop impedance is considered in {5.3.4}.

A circuit protected by an RCD will need special attention, because the earth-fault loop test will draw current from the phase which returns through the protective system. This will cause an RCD) to trip. Therefore, any RCDs must be bypassed by short circuiting connections before earth-fault loop tests are carried out. It is, of course, of the greatest importance to ensure that such connections are removed after testing. One manufacturer supplies a patented loop tester which does not require RCDs to be short circuited and which will not cause them to trip

when the earth-fault loop test is made.  Some instruments limit the test current to 15 mA so as not to trip RCDs with ratings of 30 mA and above. Whilst such tests may often be useful, they do not test the integrity of the system under fault current conditions.

When loop testing at lighting units controlled by passive infrared detectors (PIRs), there may he damage to the associated electronic switches unless they are short-circuited before testing.

An alternative to the use of a dedicated earth-fault loop impedance tester is to measure the combined resistance of the phase and protective conductors from the incoming position to the point for which earth-fault loop impedance is required (this is R1 + R2 - see {8.4.4}) and to add to it the external earth-fault loop impedance (Ze) which can be obtained from the electricity supplier. All earth-fault loop impedance test results should be carefully compared with the data in [Tables 41B and 41D], adjusted to allow for ambient temperature, or  with figures provided by the designer. To ensure that ambient temperature is taken into account, the results should never exceed three quarters of the values given in [Tables 41B and 41D].

 

 

Return to top of page

Extracted from The Electricians Guide Fifth Edition
by John Whitfield
Published by EPA Press Click Here to order your Copy

Click here for list of abbreviations