B Y R I C H A R D H U N T , M A R K A D A M I A K ,A L K I N G , & S T E E L M C C R E E R Y
THE TRADITIONAL SOLUTIONS FOR
protecting networked distribution lines,
subtransmission lines, and industrial facil-
ity incoming supply lines use some form of
pilot protection. A significant challenge for pilot protec-
tion, especially in retrofit situations, is the cost of instal-
ling pilot communications channels.
Digital radio is an inexpensive method to provide digi-
tal communications for pilot protection at the distribu-
tion level. It has the ability to send permissive, blocking,
and transfer trip signals over short to medium distances.
The relay-to-relay messaging protocols have now
become standardized through the International Electro-
technical Commission (IEC) 61850 generic object-
oriented substation event (GOOSE) profile and can
provide protection information and metering, monitor-
ing, and control as well.
This article discusses the application of pilot protection
on distribution lines. Pilot protection uses communica-
tions channels to send information between relays at each
end and is commonly used on networked lines. For the
purpose of this article, we can assume distribution lines
are circuits with an operating voltage typically between 4
and 69 kV. However, the principles discussed in this arti-
cle can be applied to any circuit at any voltage level,
assuming the protection requirements for speed, security,
and dependability can be met.
The protection challenges for pilot protection of distribu-
tion lines are identical to the protection challenges for pilot
protection on transmission lines. The major goal for pilot
protection is to operate dependably for a fault on the pro-
tected line and securely for faults outside the protected line.Digital Object Identifier 10.1109/MIAS.2009.933402
NAVY PHOTO BY JOURNALIST 2ND CLASS JACK ROUS
51
IEE
EIN
DU
STR
YA
PP
LICA
TION
SM
AG
AZIN
E�
SE
PTjO
CT
20
09�
WW
W.IE
EE
.OR
G/IA
S
Typical Applications
OverviewThe majority of the distribution sys-tems are radial systems, which allowsthe application of time-coordinatedovercurrent protection schemes. Al-though the overall distribution systemmay be designed as a radial system,individual pieces may be effectivelynetworked. The short distributionlines to industrial facilities withsignificant generating capabilities orshort lines to independent power pro-ducers may result in a small net-worked system. In addition, someparts of the distribution system areintentionally networked, such as inlarge urban load centers. In these cases,protection system must use a securemethod of identifying faults to ensureappropriate isolation of faulted sections of the system. Inany of these cases, some form of pilot protection is typi-cally applied.
Industrial Facility with Fault Current SourceAs mentioned earlier, one common example of an effec-tively networked radial distribution system is the shortdistribution line that connects an industrial facility withsome generation to a utility distribution network. Thelocal generation in the industrial facility may or may notbe large enough to carry the complete facility load, but thegeneration is a source for short-circuit current on theincoming distribution line and the utility system. Thisapplication typically requires the use of some sort of direc-tional protection or protection that can identify the faultis on the incoming distribution line. The protectionscheme must still operate correctly even if the generationis not running and there is no contribution from the plant.
Protection Solutions
OverviewThere are a variety of protection schemes for short, net-worked distribution lines. The most common methodsuse pilot communications to address concerns about coor-dination, security, and operating time. The pilot-protec-tion schemes use some form of communications betweenrelays at both line ends to ensure secure, selective tripping.
The communications medium useddepends on the protection type selected,the capabilities of the relay selected, andother factors such as cost of installation.The pilot-protection methods may sendanalog values between relays at each endof the line or use simple on/off, permis-sive, or blocking signals between relaysat each end of a line.
Pilot-Wire ProtectionThe traditional protection solution hasbeen pilot-wire relaying. Pilot-wirerelaying is an analog system using onlycurrent measurements with the relaysinterconnected between each end us-ing telephone-type copper cable.
However, the industry is movingaway from pilot-wire relaying as asolution for the protection of short,networked distribution lines. This
move has little to do with the performance of pilot-wirerelaying but has more to do with the general trend towarddigital relays and the use of digital communications. Oncedigital communications are used, current differential is abetter protection choice. Digital representation of analogcurrents can be sent between each end on a per-phase basis.
Other Common Protection MethodsOther common protection methods for the short, net-worked distribution line include line-differential relayingand pilot-protection schemes such as permissive overreach-ing transfer trip (POTT) and directional comparisonblocking (DCB) schemes. All of these methods requirecommunications between the relays at each end of the line.
However, the key piece of all of these protection meth-ods and a great challenge to reliability is the communica-tions channel. The communication channel must meetapplication requirements for performance, reliability, andcost. The performance requirements are clear: the commu-nications channel must be physically capable of sendingthe correct type of pilot signal, have enough bandwidth tohandle the signal, and have a sufficiently short systemlatency time (Table 1).
Reliability requires that the communications channelalways be available. The channel consists of the actualmedia used and any interface or conversion devices requiredbetween the relay and the communications channel. Thecost of equipment, installation, and maintenance are all
TABLE 1. COMMUNICATIONS CHANNEL REQUIREMENTS.
Protection Method Data TypeMessageSize
LatencyTime
CommunicationsMedia
Pilot-wire relaying Analog (voltage) — <1 ms Metallic pilot-wire pair
Line-differentialrelaying
Analog converted todigital message
Large 8 ms Fiberoptic
Pilot protection (POTT,DCB)
Boolean (blockingsignal, permissivesignal)
Small 8–16 ms Analog telephone line,microwave, power-line carrier, fiberoptic
DIGITAL RADIO ISAN INEXPENSIVE
METHOD TOPROVIDE DIGITALCOMMUNICATIONS
FOR PILOTPROTECTION AT
THE DISTRIBUTIONLEVEL.
52
IEE
EIN
DU
STR
YA
PP
LIC
ATI
ON
SM
AG
AZI
NE�
SE
PTjO
CT
20
09�
WW
W.I
EE
E.O
RG
/IA
S
important considerations. For a new facility or expansionproject, the cost of installing fiberoptic cable is only incre-mental. Installing fiberoptic cable or even copper cable in aretrofit installation is very expensive. The equipment costsmay wind up as a trivial part of the total project cost.
Digital RadioDigital radio is simply the use of radios to transmit digi-tal data using standard digital communications proto-cols. This is commonly used in applications rangingfrom simple voice communications to the simultaneoustransmission of multiple high-speed data channels over awide-band radio link.
The attraction of digital radio for the protection ofshort distribution lines is low installation costs, particu-larly in retrofit situations. For short lines (less than 1 miin length) with good line of sight between line ends, thetotal equipment costs can be less than US$5,000, withonly a simple engineering project for installation of theradios. The concern over digital radios is performance.The performance of other communication medium andprotection applications is well known and well understoodby protection engineers. However, digital radio is stillnew to the protection area. So, the questions of perform-ance related to the data types, bandwidth, channel latency,distance, and reliability must be understood.
Data TypesCurrently, at the physical layer, digital radios apply twocommon interface standards: RS485 and 10BaseT Ether-net. The choice of interface directly impacts the transmis-sion range and type of radio (Table 2).
Protective relaying requires the exchange of digital mes-sages between relays. Digital radios create an Ethernet net-work between the radios, which allows the use of IEC61850GOOSE messages for pilot protection.
Spread-Spectrum RadiosRadio transmission using the spread-spectrum techniquewas originally developed to provide jam-resistant militarycommunications. This technique uses a modulation tech-nique that distributes a transmitter’s signal over a very widebandwidth, making it virtually undetectable to a conven-tional radio receiver. Frequency-hopping spread spectrumand direct-sequence spread spectrum are the two primarilymethodologies the spread-spectrum technology uses totransmit messages today. The frequency-hopping spread-spectrum radios are better in environments with interfer-ence and are less likely to be jammed. The direct-sequence
spread-spectrum radios can support higher data band-widths. The spread-spectrum radios do not require an Fed-eral Communications Commission (FCC) license to installand operate.
Digital Radio Performance TestingThe best way to prove the performance of the digital radiofor pilot protection is to test the performance under fieldconditions. In this case, the test was for channel latencyand channel reliability. The protection message betweeneach relay is an IEC61850 GOOSE message (Figure 1).
Application: Pilot Protection via IEC61850 GOOSEand Spread-Spectrum RadiosThe radios were connected in a point-to-point topology,exactly as the typical pilot-protection application. Thetest protocol was a simple matter of measuring the round-trip time of each automatically generated IEC61850GOOSE message transmitted from one relay to anotherrelay and then immediately echoed back to the first relay.
This article defines channel latency as the delay betweenthe initiation of the GOOSE message by the sending relayand the reception of the GOOSE message by the receivingrelay. The test actually measures the round-trip channellatency, as one relay measures the time between sendingthe original message and receiving the echoed response. Ofinterest for pilot protection applications is the one-waychannel latency, which is assumed to be one half of thisdirectly measured round-trip channel latency.
PerformanceThe test was actually a time-based test. The IEC61850GOOSE messages were sent back and forth for somelength of time. At the end of the test, 29,672 messages
RelayRelay
GOOSE Message Path
1
~1 km
Digital radio pilot-protection test setup.
TABLE 2. DIGITAL RADIO APPLICATION GUIDELINES.
Interface Protocol Bandwidth Type of RadioTypical MaximumDistances
10BaseT DNP3 via TCP/IP, Mod-Bus, IEC61850
0.5 MB Spread spectrum: nolicense required
Up to approximately20 mi (terrainpermitting)
10BaseT DNP3 via TCP/IP, Mod-Bus, IEC61850
1.0 MB Spread spectrum: nolicense required
Up to approximately15 mi (terrainpermitting)
53
IEE
EIN
DU
STR
YA
PP
LICA
TION
SM
AG
AZIN
E�
SE
PTjO
CT
20
09�
WW
W.IE
EE
.OR
G/IA
S
were sent. The test results from the29,672 consecutive messages are re-corded in Table 3.
The typical round-trip time is 20–30 ms, meaning the channel latency is10–15 ms. This is a very acceptableperformance for pilot protection ondistribution systems. Therefore, thedigital radio meets the general per-formance requirements for pilot-pro-tection applications.
Site Installation andCommissioning ProceduresThe key issue with any communica-tions channel used for protection is thereliability of the channel, in this case,specifically the reliability of the radiopath. The important factors for thedigital radio are the distance betweentransmitter and receiver, obstructionsin the line of sight between antennas,and the natural environment beneaththe path. The short-range paths (less than 1 mi) can be vis-ually evaluated. Longer distances typically require a pathstudy that predicts the signal strength, reliability, andfade margin of a proposed radio link, as follows.
Radio communication is limited to line of sight. How-ever, radio line of sight is longer than the optical line of sightbecause of the bending of the radio wave toward the surfaceof the earth. This radio horizon is typically 30% longer thanthe visual horizon. Therefore, a longer communications pathrequires taller antennas to maintain the line of sight.
Obviously, obstructions in the line of sight willimpact the performance of the digital radio, as thestrongest radio signal is communicated directly alongthe radio line of sight. As obstructions block the widthof the radio wave front, less of the signal gets through tothe antennas. Obstructions may also cause multipath
interference due to reflective or refrac-tive signals that may arrive at thereceiver out of phase with the desiredsignal. However, because of the dif-fraction of the radio wave, objects notdirectly in the line of sight can also actas obstructions. The region whereobstructions may impact the perform-ance of the radio wave is known as theFresnel zone (Figure 2).
A Fresnel zone is one of a (theoreti-cally infinite) number of concentricellipsoids of revolution that definevolumes in the radiation pattern of theradio wave. There are multiple Fresnelzones, but only the first Fresnel zone isimportant for signal strength.
In practice, 60% of the first Fresnelzone must be clear of obstructions toallow successful radio communications.The radius of the Fresnel zone at itswidest point (at the center of the radioline of sight) can be determined by
b ¼ 17:32
ffiffiffiffiffid
4f
s, (1)
where b is the radius of the Fresnel zone, d is the distancebetween transmitter and receiver, and f is the frequencytransmitted in gigahertz.
Beyond line of sight requirements and Fresnel zonerequirements, the other concern with a digital radio path isfading or the probability that the radio signal will be lostbecause of other conditions. The fade margin determinesthe allowable signal loss between the transmitter andreceiver. It is a function of system gains (transmitter power,receiver sensitivity, and antenna gain) and system losses(free space loss, losses due to earth curvature, and coaxialcable loss). Variations in the temperature and humidity ofthe atmosphere with elevation cause the signals to bendmore or less, resulting in fading at the receiver. The longerthe path, the more likely fades will occur deep, requiring agreater fade margin. The local propagation conditionsimpact the probability of signal fade as well. Generally,mountainous terrain is favorable, and tropical areas andthose near the large bodies of water are unfavorable.
One of the losses considered when determining the fademargin is the free-space loss. The free-space loss is the lossin signal strength of the radio wave passing through thefree space. It is the basic path loss of the system. The free-space loss is defined by
where f is the frequency in gigahertz and d is the distancein km.
Like any other communications channel, proper instal-lation results in desirable performance. The installation ofthe digital radio systems requires proper site selection, anevaluation of path quality, and correct selection andmounting of antennas. The following is a brief overview ofthese requirements.
2
b
d
Fresnel zone and interference.
TABLE 3. RADIO PERFORMANCE TEST RESULTS.
Round TripTime (ms)
No. ofMessages Percentage
<20 232 0.78
20–30 29,303 98.76
30–40 127 0.43
40–80 10 0.03
THE MAJORGOAL FOR PILOT
PROTECTION IS TOOPERATE
DEPENDABLY FORA FAULT ON THEPROTECTED LINEAND SECURELY
FOR FAULTSOUTSIDE THE
PROTECTED LINE.
54
IEE
EIN
DU
STR
YA
PP
LIC
ATI
ON
SM
AG
AZI
NE�
SE
PTjO
CT
20
09�
WW
W.I
EE
E.O
RG
/IA
S
Site Selection—Basic RequirementsFor optimum radio performance, the installation sites formaster and remote stations must be carefully considered.Suitable sites should provide
n protection of the radio equipment from directweather exposure
n a source of adequate and stable primary powern suitable entrances for antennan interface or other required cablingn antenna location that provides an unobstructed
transmission path in the direction of the associatedstation(s).
Antenna Selection and OrientationThe single most important item affecting the radioperformance is the antenna system. Careful attention mustbe given to this part of an installation because the perform-ance of the entire system will be compromised. The high-quality, high-gain antennas should be used at all masterand remote stations. The antennas should be specificallydesigned for use at the intended frequency of operation.
Communication antennas fall into two general catego-ries: omnidirectional and directional. An omnidirectionalantenna provides equal radiation and response in all direc-tions and is therefore appropriate for use at master sta-tions, which must communicate with an array of remotestations scattered in various directions. At remote stations,a directional antenna such as a Yagi is typically used.Directional antennas confine the transmission and recep-tion of signals to a relatively narrow lobe, allowing greatercommunication range and reducing the chances of inter-ference to and from other users outside the pattern.
Antenna Mounting ConsiderationsThe antenna manufacturer’s installation instructions mustbe strictly followed for proper operation of a directional oromnidirectional antenna. Mount the antenna in a clearlocation, as far away as possible from obstructions such asbuildings, metal objects, and dense foliage. Choose a loca-tion that provides a clear path in the direction of the asso-ciated station.
Polarization of the antenna is important. Most systemsuse a vertically polarized omnidirectional antenna at themaster station. Therefore, the remote antennas must alsobe vertically polarized (elements perpendicular to the ho-rizon). The cross polarization between stations can cause asignal loss of 20 dB or more.
Other factors that impact the performance of the digi-tal radio include the feedline used with the antenna andthe standing-wave ratio (SWR) of the antenna system.The best feedline cable to use is a low-loss cable typesuited for 900 MHz, such as Heliax. The SWR ensure aproper impedance match between the transceiver and theantenna system. The reflected power should be less than10% of the forward power.
Digital Radio and SecurityOne concern with the digital radio is communicationssecurity. The traditional communications channels forprotection are physically isolated from computer networksand therefore carry little security risk. However, the digi-tal radio signals are theoretically available to anyone with
the proper equipment. Security issues fall into the catego-ries of protection of privacy, protection from unauthorizedaccess, and protection from denial of service attacks.
The unlicensed spread-spectrum radios, such as thosesuggested for pilot protection in this article, are inher-ently secure. The spread-spectrum technology was devel-oped during World War II for the military because of itsability to reject jamming and the difficulty in intercept-ing transmission. The only practical way to intercept mes-sages is with a stolen radio. However, all software thatoperates the radio resides within the radio. Therefore, it isdifficult to retrieve and read the messages using just a sto-len radio.
In addition, the digital radios support common secu-rity standards and methods, including AES-128 or RC4encryption standards, 802.1x RADIUS authenticationtechniques to block unauthorized access, and the SimpleNetwork Management Protocol Version 3 (SNMPv3)protocol for network management to ensure messageintegrity, authentication, and encryption.
Digital Radio Pilot-Performance ComparisonWith end-to-end performance in the 10–15 ms range, thedigital radio immediately places itself in the mix of usablepilot channels in the protection and control world. Thetraditional pilot-protection channels include power-linecarrier, audio tone, and more recently digital channels viafiber (direct or multiplexed) or copper. Although not quiteas fast as the total time of direct fiber or a wide-band car-rier set (3.5–6 ms), it compares quite favorably to analogtone over an analog microwave (13–18 ms) or a digitalchannel through a modem (15–18 ms).
This article explicitly discusses the use of an IEC61850GOOSE message transmitted over Ethernet. One reason isthat the Ethernet radios are an inexpensive and easy-to-configure solution for the digital radio applications. Inaddition, IEC61850 is a nonproprietary solution, as theGOOSE is an international standard with demonstratedmultivendor interoperability. The GOOSE is configurableto communicate multiple status, analog values, and qual-ity values in a single message. The 10–15 ms messagedelivery time mentioned earlier is invariant for GOOSEpackets containing limited combinations of the aforemen-tioned data items. Also, using an IEC61850 GOOSE mes-sage over Ethernet provides impressive error-checkingcapabilities to ensure messages are received correctly. Theradios use a 16-b cyclic redundancy check (CRC), and theGOOSE message uses a 32-b CRC. This larger CRC elimi-nates the need for security counts on received messages forpermissive or blocking signals. If the CRC is validated,there is only a one in 4 billion chance that the receivedmessage is incorrect. Also, the digital radios using Ether-net can support other communications traffic than simplyprotection. By using a virtual local-area network (VLAN),the protection GOOSE messages will always have priorityover other types of traffic, and so no channel delays occur.
There are other methods to implement the digital radioas a pilot-protection communications channel. One possi-bility is to use radios that transmit physical contact closurestates, similar to traditional power-line carrier or micro-wave solutions. Another possibility is to use a proprietarypilot-protection communications protocol available from
55
IEE
EIN
DU
STR
YA
PP
LICA
TION
SM
AG
AZIN
E�
SE
PTjO
CT
20
09�
WW
W.IE
EE
.OR
G/IA
S
various relay vendors. Using a pro-prietary protocol requires relays fromthe same vendor on each end of theline and requires the radios that canactually transmit this proprietaryprotocol. Depending on the methodselected (contact closure or protocol),radios selected, and actual protocolused, the channel latency can be sig-nificantly less than that of IEC61850GOOSE messages over Ethernet.These other methods may only workwith specific models of radios andmay require the use of radios operat-ing on licensed transmission frequen-cies. These methods will not use the32-b CRC error checking availablethrough IEC61850 GOOSE messag-ing and may not implement the 16-bCRC available in Ethernet-enabled digital radios.
Pilot Protection with the Digital RadioAlthough latency through the digital radio may be a con-cern when operating in the distribution realm, a systemlatency of 10–15 ms is typically acceptable. The other con-cern may be the need for redundancy in case the radio com-munications fail for some reason. These are similar concernsto pilot protection for transmission-line applications.
The best way to discuss these performance criteria is tolook at specific examples of pilot protection using thedigital radio as applied to the distribution feeder supply-ing an industrial facility. Specifically, we will discuss a
POTT scheme, DCB scheme, reverseinterlocking scheme, and combinationPOTT/DCB scheme. This articleassumes that all protection schemelogic is performed with a microproces-sor-based relay. Obviously, traditionalhardwire-control logic can be used aswell. It is important to note that themicroprocessor-based relay treats theGOOSE message the same as any otherdigital input. The status of the specificbit from the GOOSE message is usedin the relay logic just the same as a reg-ular contact input.
POTT SchemeThis application of the POTT schemeuses definite time-directional overcur-rent or distance elements. These ele-
ments are set to see faults beyond the other end of the lineand are configured with no intentional time delay. Instan-taneous tripping is acceptable, because a permissive signalfrom the other end of the line is required to allow trip-ping. Since the relays at both ends must send a permissivesignal for the POTT scheme to operate, there must be asource of the fault current for the protected line at eachend of the line. Depending on the capabilities of the relayused for the POTT scheme, it may be possible to imple-ment a weak infeed/echo logic to account for no source onone end.
The internal relay logic for the POTT scheme is actuallyquite simple. When either the phase or neutral direction-
al overcurrent element picks up, aGOOSE message containing a per-missive flag is sent to the relay on theother end. If a GOOSE message con-taining a permissive flag is receivedfrom the other end while either of thelocal overcurrent elements is pickedup, the relay trips the local circuitbreaker. There is no intentional timedelay in this scheme. For a fault onthe protected line, the total operatingtime of this scheme will be approxi-mately 30–35 ms, ignoring thebreaker-operating time. This assumesa channel latency of 10–15 ms andapproximately 20 ms for the relay ele-ment to operate (Figure 3).
The traditional POTT schemesusing analog communications such
IEC61850 GOOSEMessaging
IEC61850 GOOSEMessaging
IEC61850 GOOSE Messaging ViaDigital Radio Permissive Flags
Trip CB Trip CB
RO
RO
A B
Radio
Ethernet
Relay
Radio
Ethernet
Relay
67 67N 67 67N
3POTT scheme using digital radio.
PermissiveGOOSE Message
Transient Block
RO Elements67–50 Forward
67–50N ForwardOR
AND
Operate Seal-in0
6 cyc
OutputPOTT Operate
POTT Transmit
Output Contact
GOOSE MessageOutput
tpkptdpo
AND
4POTT scheme logic with transient blocking shown.
DIGITAL RADIOPERFORMANCEPLACES ITSELF IN
THE MIX OFUSABLE PILOT
CHANNELS IN THEPROTECTION AND
CONTROLWORLD.
56
IEE
EIN
DU
STR
YA
PP
LIC
ATI
ON
SM
AG
AZI
NE�
SE
PTjO
CT
20
09�
WW
W.I
EE
E.O
RG
/IA
S
as a power-line carrier or microwaveuse a pickup time delay on the permis-sive receive signal in case of spurioussignal reception. There is no need toadd a pickup time delay to the permis-sive receive signal, as this is a digitalstatus contained in the GOOSE mes-sage. Also, there is no need to add asecurity count to the permissive receivesignal. Validation through the 32-bCRC of the GOOSE message ensuresthe received message is correct. For thissimple, one-line application, there isno need to add any transient blockingdelay for current reversal. However, ifparallel lines are serving the same facil-ity or tied to the same bus, then atransient blocking delay for currentreversal must be added on to the per-missive receive signal (Figure 4).
DCB SchemeThe DCB scheme also assumes the use of definite time-directional overcurrent or distance elements. This schemerequires a forward directional overcurrent or distance ele-ment looking toward the protected line for tripping and areverse directional overcurrent or distance element look-ing behind the protected line to initiate a blocking signal.The reverse directional overcurrent or distance elementthat initiates the blocking signal isset with no intentional time delay.
The directional overcurrent ordistance element that is used for trip-ping is set with a short time delay toaccount for channel delay time. Thistime delay can be set to approxi-mately four cycles to allow for themaximum message latency of 40 msplus approximately one cycle for theremote relay to initiate the blockingsignal. Therefore, the total operatingtime to protect the line will be fourcycles ignoring breaker-operatingtime (Figure 5).
It may be necessary to add a shortseal-in timer to hold the blocking sig-nal. The blocking signal is a digitalflag contained in a GOOSE message.
Since the loss of one GOOSE messagewill cause the blocking signal to dropout, this timer ensures the blocking sig-nal is maintained when an individualmessage is lost. Assuming that theblocking GOOSE message is sent in4 ms, a one-cycle time delay means thatfour consecutive GOOSE messagesmust be lost before the block is released.This delay also means the blocking sig-nal must drop out for one cycle beforetripping is permitted (Figure 6).
Reverse Interlocking SchemeEven when the distribution feeder for alarge load or industrial facility is a radialfeed, it may be desirable to implementpilot protection on the incoming distri-bution feeder. The pilot protectionshould result in faster clearing times for
faults and alleviate many coordination issues. The pilot pro-tection in this case can be a simple reverse interlocking pro-tection scheme. In a reverse interlocking scheme, bothdownstream and upstream relays use high-speed overcurrentprotection. When the downstream relay picks up for a fault,this relay sends a blocking signal to the upstream relay. Theovercurrent element on the downstream relay is set to over-reach the downstream line end, and with no intentionaltime delay. The overcurrent element on the upstream relay
IEC61850 GOOSEMessaging
IEC61850 GOOSEMessaging
IEC61850 GOOSE Messaging ViaDigital Radio Blocking Flags
67 67N
67 67N
67 67N
67 67N
RO
RO
B
B
52 52Trip CB
Radio
Ethernet
Relay
Trip CB
Radio
Ethernet
Relay
5DCB scheme using digital radio.
BlockGOOSE Message
B Elements67–50 Reverse
67–50N ReverseOR
RO Elements67–50 Forward
67–50N ForwardOR
AND
Operate Seal-in0
6 cyc
Block Seal-in
1 cyc
Channel Delay
04 cyc
OutputDCB Operate
DCB Transmit
Output Contact
GOOSE MessageOutput
0
6DCB scheme logic.
THE REGIONWHERE
OBSTRUCTIONSMAY IMPACT THEPERFORMANCEOF THE RADIO
WAVE IS KNOWNAS THE FRESNEL
ZONE.
57
IEE
EIN
DU
STR
YA
PP
LICA
TION
SM
AG
AZIN
E�
SE
PTjO
CT
20
09�
WW
W.IE
EE
.OR
G/IA
S
is set with a short time delay of four to six cycles. This delayallows the downstream relay to detect the fault and initiatethe locking signal and allows for maximum latency (40 ms)of the digital radio message (Figures 7 and 8).
The logic for the reverse interlocking scheme is differentin the relays at the source end and the load end of the line.The relay at the load end of the line, Terminal B in thisexample, simply sends a blocking signal when an overcur-rent element picks up. The relay at the source end of the line,Terminal A in this example, can only operate when overcur-rent elements are picked up, and no blocking signal is
received from the relays at Terminal A. The reverse inter-locking is in some ways a simpler form of the DCB scheme.The operating time is similar but with the total clearingtime for a fault of the protected line of approximately sixcycles ignoring breaker-operating time.
Combination of POTT/DCB SchemeFor increased reliability, one possibility is to apply boththe POTT scheme and a DCB scheme operating in parallel.The POTT scheme should operate essentially instantane-ously for fault on the protected line. The DCB scheme,
because of the need to initiate andreceive a blocking signal, has a shorttime delay of four to six cycles. There-fore, for an internal fault, the POTTlogic should trip instantaneously. Ifthe fault fails to clear, such as for abreaker failure condition, the DCBlogic will trip in six cycles. Considerthe fault conditions if the digitalradio fails. For a fault on the pro-tected line, the POTT logic will notoperate because no permissive signalis sent or received. The DCB logicwill operate, as no blocking signal issent or received. For a fault not on theprotected line, once again the POTTlogic will not operate. However, theDCB logic will operate, as no block-ing signal is received or sent. It may
IEC61850 GOOSEMessaging
IEC61850 GOOSEMessaging
IEC61850 GOOSE Messaging ViaDigital Radio Blocking Flag
50 50N 50 50N
Trip CB
RO
B
A B
Radio
Ethernet
Relay
Trip CB
Radio
Ethernet
Relay
7Reverse interlocking scheme using digital radio.
BlockGOOSE Message
B ElementsTerminal B 50 PickupTerminal B50N Pickup
OR
RO ElementsTerminal A 50 Pickup
Terminal A 50N PickupOR
AND
Operate Seal-in0
6 cyc
Block Seal-in
1 cyc
Channel Delay
06 cyc
OutputR.I. Operate
R.I. Transmit
Output Contact
GOOSE MessageOutput
0
8Reverse interlocking scheme logic.
RO Elements67–50 Forward
67–50N Forward
PermissiveGOOSE Message
OR
AND
Operate Seal-in
06 cyc
OutputPOTT Operate
POTT Transmit
Output Contact
GOOSE MessageOutput
AND
Operate Seal-in0
6 cyc
OutputDCB Operate
DCB Transmit
Output Contact
GOOSE MessageOutput
Channel Delay4 cyc
0
B Elements67–50 Reverse
67–50N ReverseOR
BlockGOOSE Message
Block Seal-in0
1 cyc
9POTT/DCB combination scheme logic.
58
IEE
EIN
DU
STR
YA
PP
LIC
ATI
ON
SM
AG
AZI
NE�
SE
PTjO
CT
20
09�
WW
W.I
EE
E.O
RG
/IA
S
be desirable to increase the time delayof the DCB logic to allow other protec-tion to clear external faults.
This combination scheme is veryattractive when the line being pro-tected is the line to an industrial facil-ity with generation. When the facilitygeneration is running, both the POTTand the DCB scheme will operate cor-rectly. However, when the generationis not running, the POTT scheme willnot operate correctly for fault on theprotected line. The directional over-current relay at the plant end of theline will not see a fault on the pro-tected line and will therefore not senda permissive signal. However, theDCB scheme will operate correctly inthis case. No blocking signal will besent when the fault is on the protectedline. However, for a fault in the plantitself, a blocking signal will be sent tothe utility end of the line (Figure 9).
Scheme ConsiderationsThese examples show that the digital radio usingIEC61850 GOOSE messages can be the communicationschannel for the two most common pilot-protectionschemes. By extension, the digital radio can be used in anypilot-protection scheme, including directional compari-son unblocking (DCU), permissive underreaching transfertrip (PUTT), and the hybrid POTT scheme. This articleuses the POTTand DCB schemes as examples to show thatthe digital radio can perform in both a permissive andblocking logic.
The POTT scheme is a very secure scheme but will failto operate on a loss of communications channel during anin-zone fault. The DCB scheme is a very dependablescheme but may operate incorrectly for an out-of-zone faultduring a loss of communications channel. These risks havealways existed, starting with power-line carrier communi-cations. In fact, the DCU scheme using frequency shiftkeying is a scheme designed around the unreliability of thepower-line carrier signal. The correctchoice of pilot-protection schemewhen using the digital radio is there-fore part of the art and science ofprotective relaying. The philosophy,experience, and application criteriawill lead to the best solution for a spe-cific situation.
Redundancy ConsiderationsAs with any other protection scheme,there are redundancy and backupconsiderations when using the digitalradio as part of a pilot-protectionscheme. The previous application of acombination POTTand DCB schemeis one such example of redundancy. Iffor any reason the digital radio chan-nel fails, the line will still trip for
internal faults. However, there couldbe a loss of security for external faults.A simple way to add redundancy is touse two separate radio paths for com-munications. In other words, simplyuse two separate radio sets. The sameGOOSE message is sent to each radio,and both GOOSE messages are re-ceived by the relay. That way, if one setof radios fails to communicate, theother set will still operate. This meth-od requires one of the radio sets to usecross-polarized antennas to preventchannel interference (Figure 10).
Other Applications
Other Protection ApplicationsObviously, pilot connection using thedigital radio can be applied in any net-work distribution line application,including a line serving independentpower producers (IPPs) and networkeddistribution lines in downtown loadcenters. However, when the digital
radio used is an Ethernet-based radio, the radio essentiallyestablishes an Ethernet network between remote devices.This allows the extension of the protection scheme ininteresting ways.
Parallel FeedersParallel feeders from a utility serve some industrial facilities.In this case, separate pilot-protection systems are requiredfor each incoming distribution line. However, with the digi-tal radio, one set of radios can be the communication channelfor multiple sets of lines. Each relay simply sends a GOOSEmessage to the radio or receives a GOOSE message from theradio as appropriate. Using two sets of radios in this caseprovides complete reliability of communications. Eachpilot-protection system sends its tripping or blocking sig-nals over both sets of radios. Therefore, both lines have aprimary and secondary communications channel whilerequiring only two sets of radios (Figure 11). To extend this
IEC61850GOOSE
Messaging
IEC61850GOOSE
MessagingIEC61850 GOOSE Messaging ViaDigital Radio Permissive Flags
67 67N67 67N
Trip CB Trip CB
RO
RO
A B
Ethernet
Relay
Ethernet
Relay
Radio
Radio Radio
Radio
10Redundant digital radio application.
THE INSTALLATIONOF THE DIGITALRADIO SYSTEMS
REQUIRES PROPERSITE SELECTION,AN EVALUATION
OF PATH QUALITY,AND CORRECTSELECTION ANDMOUNTING OF
ANTENNAS.
59
IEE
EIN
DU
STR
YA
PP
LICA
TION
SM
AG
AZIN
E�
SE
PTjO
CT
20
09�
WW
W.IE
EE
.OR
G/IA
S
example even further, consider a plant that has three incom-ing utility feeds. Two sets of radios provide a redundantcommunications path for all three incoming feeders.
Other Digital RadioChannel ApplicationsGiven a digital communication channel in a substation, awide variety of ancillary functions becomes available.When using a digital radio channel in an application, thechannel is available 99.9% of the time for other applica-tions. In the case of an Ethernet-based radio, the Ethernetcan be connected to all other devices in the remote substa-tion and provide complete data access. This access wouldtypically include supervisory control and data acquisition(SCADA) (with remote control), remote setting/softwareupdates, oscillography and sequence of events retrieval,and physical security monitoring.
Besides the typical substation functions mentionedearlier, digital connectivity enables the ability to transmitdigital images. Specifically, many remote control functionsrequire visual confirmation of an operation such as theopening of a disconnect. By providing a position-selectablecamera, an operator can position the camera to focus on asubstation device (e.g., switch), visually check the status ofthe device before the control operation, execute the controloperation, and then verify the result of the operation.
Another recently field-tested application had to do withmobile verification of the angle reference of distributionfeeders. In this application, there was an operational need tobe able to verify the phasing between a substation source andthe service in a customer location. The distance between thesubstation and the customer premises could range from doz-ens of meters to 2 km. The solution of this mobile-monitor-ing application was the use of a set of phasor measurementunits (PMUs), synchronized by a set of global positioningsystem (GPS) clocks, and communicating with GOOSEthrough the Ethernet digital radios. Each PMU measuredthe absolute local angle (either the source angle or the
customer’s service angle) and each endthen communicated the measured angleto the other via GOOSE. The receivedangle was subtracted from the locallymeasured absolute angle, and the rela-tive difference was then displayed.
SummaryThis article describes the basic digi-tal radio technology and shows somepossible applications of the digitalradio for distribution protection. Aswith any other pieces of the protec-tion system, it is important to under-stand the reliability and performanceof the digital radio. The test resultsdocumented in this article show thatthe digital radio successfully sends anIEC61850 GOOSE message within10–15 ms 99% of the time. In nocase during the test was a messagenot received. Therefore, the digitalradio is reliable enough to use as partof the distribution protection system.
The 10–15 ms channel latency is more than acceptable fordistribution protection, is interoperable, and requires nospecial adaptation of standard pilot-protection schemes.The channel latency compares quite favorably to that ofanalog tone over an analog microwave or a digital channelthrough a modem, meaning that the digital radio isappropriate for pilot protection communication on sub-transmission lines.
The advantages of the digital radio in being able to estab-lish an Ethernet network were presented. Additionally, thedigital radios support any standard protocol over Ethernet,including Modbus, DNP 3.0, and IEC61850. At a mini-mum, this allows the digital radio to send the binary signalsnecessary for pilot protection, such as permissive and blockingsignals. In addition, the digital radio, like any other Ethernetnetwork, allows simultaneous traffic. Therefore, the digitalradio can support communications for pilot protection,SCADA communications, and metering communicationssimultaneously without any degradation in performance.
When one looks at the capabilities of the digital radioand the low installed cost of digital radio, many interest-ing applications present themselves.
References[1] J. L. Blackburn, Protective Relaying Principles and Applications, 2nd ed.
New York: Marcel Dekker, 1998.
[2] J. R. Fairman, K. Zimmeramn, J. W. Gregory, and J. K. Niemira,
‘‘International drive distribution automation and protection,’’
presented at the 27th Annual Western Protective Relay Conf., Spo-
kane, WA, Oct. 24–26, 2000.
[3] IEEE Guide for Protective Relay Applications to Transmission Lines, IEEE
Standard C37.113-1999.
Richard Hunt ([email protected]), Mark Adamiak, AlKing, and Steel McCreery are with GE Digital Energy inMarkham, Ontario, Canada. Hunt is a Senior Member of theIEEE. Adamiak is a Fellow of the IEEE. This article firstappeared as ‘‘Application of Digital Radio for DistributionPilot Protection’’ at the 2008 Rural Electric Power Conference.
