Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

PERMANENT CONNECTIONS FOR SUBSTATION EARTHING

Curtis R. Stidham and W. Keith Switzer


ERICO, Inc.
United States of America

Abstract

Proper substation earthing is required to insure personnel safety, both for the utility worker and the general
public, as well as to provide protection for equipment and minimize interruption of service. Essential to the
earthing system are the many earthing connections used to interconnect the earthing conductors and for the
bonding of structures within the system. Earthing connections must be permanent; they must be able to last the
lifetime of the installation without corroding or increasing in resistance. Two independent test programs were
sponsored, at different laboratories; to evaluate the performance of commonly used connections for substation
earthing. One program conducted was a complete test to the requirements of IEEE 837-1989 “IEEE Standard
for Qualifying Permanent Connections Used in Substation Grounding” [1]. The second program was a current-
thermal cycle test, modified from the IEEE 837-1989 current-temperature cycling test to include aged
conductor, extended duration, and a higher current level. This paper presents the results from these
commissioned tests. Results show definitively that the performance of mechanical and compression earth
connectors are not equivalent to exothermically welded connections. Results also show that only exothermically
welded connections passed the complete requirements of IEEE Std 837-1989.

1. INTRODUCTION This resulted in the publication of IEEE Std. 837-


1984, “IEEE Standard for Qualifying Permanent
Historically, many studies have been made to Connections Used in Substation Grounding” which
determine the fusing currents of various materials was later revised in 1989. The 1986 revision of
used for earth conductors and of the connectors used ANSI/IEEE Std. 80 recognized this standard and
to interconnect the conductors. It was found that allowed any connector that passed the requirements of
connectors used for substation earthing, originally IEEE Std. 837 to be used without temperature de-
designed for power applications, would not carry as rating the earth conductors. The impending 1999
much current as the conductor under fault conditions. revision of IEEE Std 80 will recommend that only
Substation earthing criteria was first published in the connections that pass the requirements of IEEE Std.
USA in AIEE No. 80, March 1961, “Guide for Safety 837 be used in substation earthing.
in Alternating-Current Substation Grounding”, later
changed to IEEE No. 80. The current revision (1986) This paper presents the results from two
of the guide, IEEE Standard No. 80, “IEEE Guide for independently conducted test programs sponsored by
Safety in AC Substation Grounding” [2], further ERICO, Inc. The programs were conducted to
refines the connection requirements by recognizing evaluate the ability of commonly used earthing
that exothermic connections have the same fault connectors to meet the requirements of permanent
current and thermal capacity as the copper earthing connections for substation earthing systems.
conductor. Exothermic connections were not
previously mentioned although some users recognized 2. APPROACH
their capacity, while others classified them as brazed
connections. Earthing conductors were de-rated to In order to provide reliable, impartial results, to the
250o or 350o C when mechanical (including standards committee and the electrical utility
compression) connectors were used. community, ERICO, Inc. commissioned the
laboratories of two different power utilities in North
As the 1986 revision of ANSI/IEEE Std 80 was being America to conduct testing as specified under
prepared, several manufacturers of connectors on the contract. In 1996, Ontario Hydro Technologies
revision task force requested that a new standard be (OHT), Ontario Hydro Power, of Toronto Canada was
prepared to test and qualify mechanical connectors commissioned to conduct a complete IEEE Std. 837
for permanent high fault current usage in substations. test program. The program was conducted and the

1
final test report completed “Substation Grounding
Table 1: Connectors Tested by OHT
Connectors IEEE STD 837-1989 Test Series” [3].
Also in 1996, the Southern Electric International’s Manufacturer Sample ID Part Number
(SEI) Georgia Power Research Center was ERICO, Inc. “CW” C-C XAC2Q2Q
commissioned to conduct a current-thermal cycling Cadweld
test modified to be more stringent than the current– exothermically
temperature segment of the IEEE Std 837-1989. The welded “CW” C-R GTC182Q
program was conducted and the final report connection
completed “Comparative Grounding Connector Test” Burndy Electrical “B” C-C YGL29C29
[4]. This paper presents the results from these die compression
commissioned test programs. The approach and test “B” C-R YGLR29C34
results from each program are presented separately.
AMP, Inc. “A” C-C 812281-1
bolted wedge
2.1 OHT Test - IEEE Std 837-1989
connector “A” C-R 81229-1
Exothermically welded connections from the
requesting company, along with mechanical and Thomas & Betts “T” C-C GG40250-
compression connectors of three manufacturers who die compression 40250
claimed their connectors met the requirements of “T” C-R GG500-40250
IEEE Std. 837 were tested. Two types of connectors C-C is conductor to conductor connection.
were tested; cable to cable and cable to ground rod. C-R is conductor to ground rod connection.
The cable used was 4/0 AWG, 19 strand, medium,
hard drawn copper. The ground rods were 17.3mm 2.1.1 Mechanical Pullout
(¾ in) diameter copper-bonded steel. OHT purchased
all materials on the open market and installed all of In the mechanical pullout test, a sample of at least
the connectors and exothermically welded 25.4cm is placed in a tensile testing machine and a
connections following the manufacturers’ load applied with the cross-head rate not to exceed
recommendations. The testing was conducted to the 2.1cm/minute/meter of sample length. No visible
requirements as prescribed in IEEE Std 837-1989. movement of the pre-marked conductor is allowed at
the specified minimal tensile load. The minimal load
IEEE Std 837-1989 requires that a minimum of four for cable to cable connections is 2225 N and 4450 N
identical connector specimens be tested in each of the for the cable to ground rod connections. A Satec
four qualification test categories. Four connector Universal machine model 120WHVL was used for the
samples of each manufacturer were used in each test test.
segment of the program. The four test categories are
mechanical pullout, electromagnetic force, acidic 2.1.2 Electromechanical Force
sequential test series, and alkaline sequential test
series. The failure of one specimen in any one of the The electromechanical force test consists of a test
four test series constitutes the failure of the connector loop, constructed of an individual manufacturer’s
to meet the requirements of the standard. The connectors, specified conductor(s), equalizers, and
connectors included in the test program are listed in includes a segment of control conductor used as a
Table 1. A general performance criteria is that the reference for resistance. Equalizers are integral to the
initial resistance of any installed connector, measured loop and are located mid-distance between each
between equalizers, shall not be greater than 110% of connector to provide resistance measurement points.
an equal length of control conductor before any tests The distance between equalizers is equal to the length
are initiated. The only exception is for the mechanical of the control conductor. The loop is subjected to an
pullout test, where resistance is not measured. The asymmetric current that is calculated to have a peak
resistance of a connection is calculated as the value of the first half cycle equal to 2.7 times one half
measured resistance between equalizers minus the of 80% of the RMS symmetrical fusing current of the
measured resistance of the control conductor. All conductor. The 80% RMS value is limited to a
resistance measurements were made with a Valhalla maximum of 100kA. An asymmetric current with a
micro-ohm meter model 4300B. peak value of 72.9 kA was used. The duration of the
surge is 0.2 seconds and the surge is applied three
times with the control conductor being allowed to
cool to a maximum 100o C before a subsequent surge

2
is applied. The power source (60Hz) was three single- thaw cycles. A cycle consists of two hours at –10o C
phase transformers connected appropriately to attain or lower and two hours at 20o C or higher. The
the required test level. Current limiting reactors were passing criterion calls for a maximum increase in
used to control the test current and provide a high resistance of 150% of the initial connector resistance.
X/R ratio. The passing criterion calls for no visible The freeze-thaw tests were performed in a large walk-
movement of the pre-marked conductor and the in environmental test room having a controlled
maximum resistance increase is 50% of the initial temperature range of –50o C to 50o C. Specimens
resistance value. were thoroughly dried before final resistance
measurements were made.
2.1.3 Sequential Test Series
2.1.3.3 “Below Grade” Acidic Corrosion Test
The acidic and alkaline sequential tests series are
conducted in the same manner expect for the This portion of the sequential test is for connectors
corrosion segment of the series. The sequence of designed for direct burial application. Half of the test
testing is as follows: current-temperature cycling, loops were subjected to this segment of the sequential
freeze-thaw, corrosion (acidic or alkaline) segment, series, while the other half were subjected to the
and fault current testing. Each series test loop used “Above Grade” alkaline test segment, described in the
though out the sequential test series consists of a loop following section. A 10% by volume nitric acid
constructed from an individual manufacturer’s solution is used for this corrosion test. Each connector
connectors, specified conductor(s), equalizers, and and 80% of the conductor between equalizers are
includes a segment of control conductor used as a submerged in the acid for a duration of time that will
reference for temperature and resistance. Equalizers result in a decrease of 20% in the cross sectional area
are integral to the loop and are located mid-distance of the control conductor when copper conductors are
between each connector to provide resistance used. When copper-bonded conductors are used, the
measurement points. The distance between equalizers duration is defined as the time required to increase
is equal to the length of the control conductor. Each the resistance of the control conductor by 25%. The
connector is instrumented with a thermocouple. passing criterion calls for a maximum increase in
Separate test loops are used for each acidic and resistance of 150% of the initial connector resistance.
alkaline series. Specimens were washed and thoroughly dried before
final resistance measurements were made.
2.1.3.1 Current-Temperature Cycling
2.1.3.4 “Above Grade” Alkaline Corrosion Test
The test loop is subjected to alternating current of a
level sufficient to raise the control conductor This portion of the sequential test is for connectors
temperature to a prescribed temperature. A designed for above grade application. The test loop is
temperature of 350o C was used for both the cable to subjected to salt fog for a minimal period of 500
cable and the cable to ground rod connections. The hours. The test is conducted per ANSI/ASTM B117-
test consists of 25 cycles, one-hour at 350o C 85 [5]. An Industrial Filter & Pump Mfg Co salt-fog
temperature. The loop is allowed to cool to ambient corrosion test chamber was used to conductor the test.
between cycles. The nominal current for the cable to The passing criterion calls for a maximum increase in
cable connections was 1010A and the cable to ground resistance of 150% of the initial connector resistance.
rod was 540A. The resistance and temperature are The specimens were washed and thoroughly dried
measured at the beginning of the first cycle and after before final resistance measurements were made.
every fifth cycle. The passing criterion call for a
maximum increase in resistance of 150% of the initial 2.1.3.5 Fault-Current Test
connector resistance and the temperature of a
connection can not exceed the temperature of the The fault-current test is conducted on all of the test
control conductor. loops as the final segment of sequential testing. The
objective of this final test is to determine if the
2.1.3.2 Freeze Thaw Test connections that have been conditioned by previous
segments of the test program can withstand fault-
The test loops from the current-temperature cycling current surges. The symmetrical fault current is 90%
are next subjected to the freeze-thaw test. The test of the fusion current for the remaining cross-sectional
consists of submerging the loops in a minimum of area of the control conductor. Three surges are
2.54cm of water and subjecting them to 10 freeze- applied for a duration of 10 seconds each. The

3
control conductor is allowed to cool to 100o C segment of control conductor (1.22m), used as a
between each surge. The passing criterion calls for a reference for temperature and resistance. The
maximum increase in resistance of 150% of the initial equalizers are integral to the loop and are located
connector resistance. mid-distance between each connector to provide
resistance measurement points. The distance between
2.2 SEI Comparative Grounding Connector Test equalizers was equal to the length of the control
conductor.
Southern Electric International’s (SEI) Georgia
Power Research Center was commissioned to conduct The test was conducted at currents (60Hz) higher
a current-thermal cycling test to evaluate the relative than specified in IEEE Std. 837-1989. The current-
performance of permanent earth connectors installed thermal cycling tests consisted of 100 cycles at a
on reclaimed cable, under sustained fault conditions. control conductor temperature of 550o C (51% of
It is common that connections must be made to an conductor fusing). The temperature specified in IEEE
existing earth system (reclaimed cable), this situation Std 837-1989, for copper conductor, is 350o C or 32%
is not addressed in IEEE Std 837-1989, therefore, of the conductor fusing temperature. Upon the failure
that condition was simulated in this test program. of a connector or conductor, it was shunted with a
segment of conductor brazed between the equalizers,
Exothermically welded connections manufactured by thus bypassing the failed section of the loop and thus
the requesting company and mechanical and permitting the test to be continued.
compression connectors from the same three
manufacturers that were tested by OHT in the IEEE Each connector and the control conductor were
Std. 837-1989 tests, were evaluated. The connectors instrumented with two thermocouples. The
tested were for cable to cable connections. SEI temperature readings from each thermocouple
purchased all materials on the open market and installed on the connectors, were record at the end of
installed all of the connectors and exothermically each cycle. The resistance was measured and recorded
welded connections, following the manufacturers’ at the beginning of the test and at every 10th cycle.
recommendations. The list of manufacturers and part The resistance of the connection is calculated as the
numbers included in the program are listed in Table measured resistance between equalizers minus the
2. measured resistance of the control conductor. The
resistance values were corrected to 20o C.
Table 2: Connectors Tested by SEI
Manufacturer Sample ID Part Number The passing criteria were the same as those
ERICO, Inc. CW1, CW2 TAC2V2V established in the current-temperature cycling test
Cadweld segment of the IEEE Std.837-1989 test. The
exothermically temperature of the connector must not exceed that of
welded tee style the control conductor. The resistance of the connector
connection must not exceed 150% that of its initial value. All
Burndy die BC1, BC2 YGL29C29 cycles were video taped.
compression,
cross connector 3. RESULTS AND DISCUSSION
Burndy die BL1, BL2 YGHC29C29
compression, tap The results from each program are presented
connector separately, and were as follows.
AMP, Inc. bolted A1, A2 81228-1
wedge connector 3.1 OHT Test - IEEE Std 837-1989
Thomas & Betts TB1, TB2 GG40250-
The OHT test results are summarized in Table 3. The
die compression, 40250
results show that only the Cadweld exothermically
cross connector
welded connections met the complete test
requirements of IEEE Std. 837-1989. There was a
The conductor, 250 kcmil (127 mm2), 37 strand, soft change in the method used to calculate the change in
drawn copper, was artificially aged by heating in an connector resistance when the conductor was a
oven at 260o C for a period of 24 hours, to simulate copper-bonded, steel cored ground rod. The resistance
oxidized conductors found in an existing earth grid is normally determined as discussed in Section 2.1.
system. The series test loop consisted of aged cable, The problem observed was that the copper thickness
two connectors for each type tested, equalizers, and a

4
varied on the ground rods and the thickness on the 3.1.2 Electromagnetic Force
control conductor could vary from that on the rods
used in the test loops. It was difficult to subtract the Failures resulted in both types of connections, cable to
resistance of the control conductor from the connector cable and cable to ground rod. The failures in the
specimen resistance and get a meaningful connector cable to cable connections were three of the four “A”
specimen resistance. It was determined that if an 8% C-C connectors displaying movement up to 0.64cm.
to 10% change in resistance was observed, within the The failures on the cable to ground rod connections
overall connector specimen resistance (connector and were three of four “B” C-R and three of four “T” C-R
conductors), this corresponded to a 150% change in connectors each displaying movement, 0.16cm and
the resistance of the connector itself. This is because 7.62cm respectively. The conductor pulled completely
the resistance of just the connector is only a small out of one of the “T” C-R connectors on the third
percentage of the overall connector specimen surge.
resistance. Therefore, the investigators changed the
definition for the criteria of allowable resistance 3.1.3 Acidic Sequential Test Series
variation, when copper-bonded ground rods were
used. The new criteria called for a maximum, overall The only manufacturer to have both styles of
connector specimen resistance of 10 percentage points connectors, cable to cable and cable to ground rod,
more than that of the control conductor. For example, pass this test series were the Cadweld exothermically
a 4% change on the control conductor meant that the welded connections. The “A” C-R connectors and the
maximum allowable increase in the overall connector “B” C-C connectors also passed this test series. The
specimen resistance would be 14%. This criterion failures of the other connections occurred at various
was applied to all connectors installed on copper- points throughout the test series. All the connectors
bonded ground rod conductors. that failed met the failure criteria of an increase over
150% in resistance.
3.1.1 Mechanical Pullout
All of the “A” C-C connector specimens failed this
The only failures in this test segment where on cable sequence test as a result of high resistance. All failed
to ground rod connections. The failures were with one after the freeze-thaw series (1417%, 351%, 1481%,
“A” C-R connector specimen and all four of the “T” and 408%). All of the “A” C-R connectors passed this
connector specimens, with the ground rods pulling test sequence.
out of the connectors below the 4450N load level. The
“A” C-R connector failed just below the minimum All of the “B” C-C connector specimens passed this
level at 4317 N and the four “T” C-R connectors sequential test. Two of the four “B” C-R connector
failed at 966, 1633, 1455, and 1108 N, well below the specimens failed this series as a result of high
4450N level. resistance. All of the “T” C-C failed this sequence
test as a result of high resistance. All of the
Table 3. Summarized results from OHT tests. specimens exceeded the resistance limit following the
Sample ID Mech- Elect- Sequential Tests freeze-thaw series (473%, 2590%, 841%, and
Pullout Mech 1139%). One of the “T” C-R connector specimens
Test Test exceeded the resistance limit following the fault
Acid Alkaline current test.
“A” C-C 4 Pass 3 Fail 4 Fail 4 Fail
All of the “CW” C-C and C-R connections passed
“A” C-R 1 Fail 4 Pass 4 3 Fail
this series of testing. The “CW” connections had the
Pass
lowest initial resistance values of all the connectors.
“B” C-C 4 Pass 4 Pass 4 4 Fail
The largest increase in resistance, upon remove from
Pass
the acid bath, was 49% on one of the “CW” C-C
“B” C-R 4 Pass 3 Fail 2 Fail 3 Fail specimens. The resistance dropped to 21% on this
“T” C-C 4 Pass 4 Pass 4 Fail 4 Fail connection after fault current testing.
“T” C-R 4 Fail 3 Fail 1 Fail 4 Fail
“CW” C-C 4 Pass 4 Pass 4 4 Pass 3.1.4 Alkaline Sequential Test Series
Pass The only connectors to pass this test series were the
“CW” C-R 4 Pass 4 Pass 4 4 Pass Cadweld exothermically welded connections. Both
Pass types of connectors, from all the remaining
manufacturers, failed during this test sequence. All

5
the connectors that failed the test either fused open or temperatures below that of the control conductor. The
met the failure criteria of an increase over 150% in time to failure for each of the test connectors is shown
resistance. in Figure 1.

All of the “A” C-C connector specimens failed this


series. One connector specimen was lost (failed open) CW1 100
during the current-temperature cycling. Two CW2 64
specimens failed as a result of resistance increase BC2 33
(467 and 346%), following the current-temperature TB1 20
cycling series. Arcing was observed during the fault

CONNECTOR
TB2 19
testing on all the specimens and one of the specimens BL2 17
failed open during the third fault cycle. The sample BC1 16
that failed open during the fault cycling had an BL1 13
increase of 475% after the freeze-thaw segment of the A1 6
testing. Three of the four “A” C-R connector
A2 4
specimens failed following the fault current testing,
0 20 40 60 80 100
as a result of high resistance. CYCLE

Figure 1: Cycles Completed


All of the “B” C-C connector specimens failed this
series as a result of high resistance. One specimen
The mode of failure for connectors “A” and “BL”
failed after the freeze-thaw series (181%). One
appeared to be the development of high contact
specimen failed following the salt-spray series
resistance between the connector and the conductor,
(154%), and the other two failed following the fault
as noted by the cherry red color and arcing observed
current test (185% and 383%). Three of the four “B”
within the connectors before failure.
C-R connector specimens failed due to high resistance
following the fault current test. Arcing was observed
Failure of connectors “BC” and “TB” occurred due to
on one of the “B” C-R specimens during the first
unequal current sharing among conductor strands and
cycle of the fault current test.
resulted in fusing of those strands. This resulted in
current overloading of the remaining strands, and
All of the “T” C-C and all of the “T” C-R connector
subsequently, the conductor fused adjacent to the
specimens failed this test sequence. All of the “T” C-
connector. The typical scenario is for the outer
C specimens exceeded the resistance limit following
strands to fuse since a higher current density exists in
the freeze-thaw series (1570%, 1668%, 526%, and
them as compared to the inner strands. This results
1394%). One of the “T” C-C connector specimens
from the high resistance that exists between the
fused during the first fault current cycle. Three of the
strands due to oxide layers and the fact that the
“T” C-R specimens exceeded the resistance limit
current must pass through the outer strands, which
following the salt-spray test. The fourth exceeded the
are the only strands in contract with the mechanical
resistance limit following the fault current test.
or compression connector. This does not occur with
the exothermically welded connections because the
All of the “CW” C-C and C-R connections passed
welding process fuses the ends of each strand
this series of testing. The “CW” connections had the
together, resulting in even current distribution among
lowest initial resistance values of all the connectors.
the strands.
The largest increase in resistance, by the completion
of the fault current testing, was 7% on one of the
As the test progressed, the conductors in the loop
“CW” C-C specimens.
started to deteriorate. A heavy oxide layer developed
and began to rain off of the conductor. The effect is
3.2 SEI Comparative Grounding Connector Test evident by the steady increase in the control
conductor resistance. Figure 2 plots the resistance of
By the completion of the first current-thermal cycle, each connector and the control conductor to failure.
the temperatures of all the connectors, except the The conductor in the section containing an
exothermic connections, had exceeded the exothermic connection CW2 failed midway between
temperature of the control conductor. By the end of the connection and the equalizer and was attributed to
the 33rd cycle, only the two exothermic connections the failure of the conductor and not caused by the
(CW1, CW2) had not failed due to fusing. At that connector.
point, both of the exothermic connections had

6
250 4. CONCLUSIONS

Results show definitively that the performance of


200
CW1 mechanical and compression earth connectors are not
CW2
TB1 equivalent to exothermically welded connections.
RESISTANCE (MICROHMS)

TB2
150
BC1
BC2
Results also show that only exothermically welded
BL1
BL2
connections passed the complete requirements of
100 A1
A2
IEEE Std 837-1989. The results from both programs
CONTROL
indicate that only the exothermic connections provide
50 a permanent and reliable connection for substation
earthing. Based on this test, only the exothermic
0 connections tested can be relied on to carry the high
0 10 20 30 40 50 60 70 80 90 100 current faults common in electrical power substations.
CYCLES

Figure 2: Connector/Control Conductor Resistance 5. REFERENCES

Figure 3 shows the average temperatures at each [1] IEEE Std 837-1989, “IEEE Standard for
cycle, until failure, of each connector specimen and Qualifying Permanent Connections Used in
the control conductor. Evident is the fact that the Substation Grounding”, The Institute of Electrical
temperature of all but the CW connections exceeded and Electronics Engineers, Inc., 345 East 47th Street,
the control conductor temperature, by the end of the New York, NY 10017, USA, 1989.
first cycle.
[2] ANSI/IEEE Std 80-1986, “IEEE Guide for Safety
in AC Substation Grounding”, The Institute of
900 Electrical and Electronics Engineers, Inc., 345 East
850
Control
CW1
47th Street, New York, NY 10017, USA, 1986.
CW2
800 TB1
TB2

750
BC1
BC2
[3] Maurice C.G., “Substation Grounding Connectors
700
BL1
BL2 IEEE Std 837-1989 Test Series”, Report No. C-95-
TEMP (C)

A1

650
A2 EST-193-P, Ontario Hydro Technologies, Electrical
600
Systems Technology, 1996.
550

500
[4] McKoon, T.L., “Comparative Grounding
450
Connector Test”, Project No. C94901, Southern
Electric International, 1996.
400
0 10 20 30 40 50 60 70 80 90 100
CYCLES
[5] ASTM B117-85, “Method of Salt Spray (Fog)
Figure 3: Connector/Conductor Temperature Testing, American Society of Testing Materials, 1916
Race St, Philadelphia, PA 19103, USA, 1985.

You might also like