Guide to electricity network design and planning – part 3
This is part 3 of a three-part guide aimed at qualified engineers and/or graduates who are beginning their career in an electricity distribution network planning department, or perhaps moving into a planning role after some years in a different role (for example operations, maintenance or construction).
Author(s): Dave Openshaw
The focus if this article is on distribution rather than transmission planning. Nevertheless, the general principles are essentially similar, and since the distinction between the two is becoming increasingly blurred given the proliferation of distributed generation and energy resources, planning engineers new to either discipline should find the three parts of this guide useful. Part 3 – which covers the basics of protection and overall protection schemes.
The two complementary parts of this guide are:
- Part 1 – which covers the general principles of network design and planning, security of supply, network design basics, voltage management, load forecasting, electrical losses, and analysis and modelling. It includes a useful 21-point check list which planning engineers may find useful when preparing proposals for network investment; and
- Part 2 - which covers legislation and regulation, quality of supply, engineering recommendations and technical reports, investment appraisal, low carbon transition, and innovation and smart grids.
Basics of electrical protection
The planning and design of electrical protection schemes for distribution and transmission networks is a specialist area of power system engineering, and many comprehensive books have been written on the subject. Transmission and distribution companies will have in-house specialists for designing, commissioning and routine testing of protection, and planning engineers should always access that expertise when planning new networks, extensions to networks, or replacements of existing assets.
Determining protection settings for any new switchboard, or indeed an existing switchboard where a change of network configuration or loading is planned, is a particularly specialist area. The need is to establish settings that provide the requisite discrimination and co-ordination under intact and outage conditions – i.e. when fault levels (and hence prospective short-circuit currents) will be significantly reduced.
The increasing proliferation of DG and electrical energy storage, along with future more variable fault levels, serves to exacerbate the challenges in designing protection systems that are able to reliably provide the required level of discrimination, coordination and sensitivity. The drive to improve quality of supply performance places further challenges on the design of protection schemes to minimise the extent and duration of supply interruptions in the event of system faults.
Notwithstanding quality of supply incentives, the primary driver for the design of reliable protection schemes is that of public and employee safety, and the prevention of damage to electrical plant.
Electricity Safety Quality and Continuity Regulation 6 requires that protection will, so far as is reasonably practical, prevent any current from flowing in any part of the network for such period of time that that part of the network can no longer carry that current without danger. Moreover, Regulation 23(b) requires that the network is provided where necessary with fuses or automatic switching devices appropriately located and set as to restrict, as far as is reasonably practicable, the number of consumers affected by a fault. In other words, there must be sufficient zones of protection, with adequate discrimination (see definition below), to limit the amount of network de-energised by the operation of protective devices.
By the same token, Electricity at Work Regulation 11 requires that efficient means, suitably located, shall be provided for protecting from excess current every part of a system as may be necessary to prevent danger, whilst Regulation 12 requires that where necessary to prevent danger, suitable means shall be available for cutting off supply to any electrical equipment and the isolation of any electrical equipment.
The following are some of the more common forms of electrical protection that will be encountered on transmission and distribution systems, together with a very brief description of their more common applications. However, a few of the fundamental principles of electrical protection and the terms used should first be explained.
Discrimination – the ability of an overall protection system (involving more than one individual protection device) to limit the extent of a circuit outage following a fault, by ensuring that the nearest protection device upstream of the fault operates prior to any other protection device designed to clear a similar fault.
Coordination – the ability of a protection system to follow a sequence of operation in order either to limit the extent of any circuit outage following a fault, or to ensure restoration of supplies to any part of the network interrupted once the protection system has isolated the fault. This is a basic requirement for automation and auto-reclose systems.
Auto-reclose – a protection scheme that enables the automatic re-energisation of a system following a fault clearance – including where appropriate, presenting a different protection characteristic on re-energisation. For overhead interconnected transmission circuits, delayed auto reclose schemes are generally used in order to allow any power swings which might occur on breaking an interconnection to decay before an auto-reclose attempt is made (imposed delays can be upwards of 5 s and up to 60 s).
Dead time – the set period between auto-reclose attempts. Auto-reclose schemes will often be set to attempt a limited number of reclosures after a short delay, followed by a final attempted reclosure after a longer delay. The ‘short’ delay will be set so as to give adequate time for a transient fault to clear and the surrounding air to de-ionise so as to improve the chances of a successful reclosure.
Reset time – a period of time following operation of a protection system necessary for the protection system to resume its normal state.
Reclaim time – the period of time following operation of an auto-reclose sequence before the system is ready to undertake a further full auto-reclose sequence.
Pick-up and drop-out – the values of the CT secondary current passing through a protection relay at which the protection will switch on or begin to operate, and switch off or cease to operate. It is important to ensure a setting that enables the relay to reliably detect fault current but remain stable when the circuit is passing a high level of load current.
Sensitivity and stability – terms relevant to unit or balanced differential protection schemes which respectively describe the ability of the protection to operate for high impedance (low current) faults within the protection zone, and to not operate for faults outside the protection zone.
Back-up protection – protection applied to a system that will operate in the event of failure of the primary protection, or if the primary protection is unable to clear the fault (it follows that the back-up protection must be able to ‘discriminate’ with the primary protection, though not necessarily with other back up protection).
Unit protection – a protection scheme that will operate only for faults within a defined zone of the network (the advantage being that since no time-delay is necessary in order to provide discrimination with any downstream protection device, the protection can operate ‘instantaneously’). The term ‘balanced’ or ‘differential’ protection is often used to describe a unit protection scheme.
Intertripping – the means by which the operation of a protection device at one location will initiate the tripping of a protection device at another location. This is generally achieved through pilot circuits. For example, in the event of a fault at a primary substation on the low voltage side of the 33/11 kV transformer (such as the 11 kV busbars) or within the transformer itself, it is necessary to ensure that the fault is cleared by tripping the 33 kV transformer feeder circuit breakers at the upstream grid (132/33 kV) substation, the protection for which will not necessarily detect the fault.
Intertripping may also be used to disconnect distributed generators in the event of a network fault clearance leading to an unacceptable system fault level.
Fault throwing switch - not a protection device as such, but for overhead line transformer feeders where pilot cables and therefore intertripping may not be practicable (wireless forms of communication traditionally being regarded as too slow or unreliable) a 33 kV fault throwing switch may be connected to the incoming 33 kV feeders at the primary substation.
Closure of the FTS (initiated by the operation of the main tripping relay at the primary substation) has the effect of earthing one phase of the live incoming 33 kV transformer feeder in order to initiate operation of the upstream 33 kV feeder earth fault protection, so clearing the fault at the primary substation.
Distance protection – the primary form of protection applied to an interconnected transmission system, allowing fast operation and back-up whilst maintaining adequate discrimination. Correct coordination between distance relays is achieved by controlling the ‘reach’ and tripping times of the various zones.
A conventional distance protection scheme will typically comprise an instantaneous directional element and one or more time delayed elements referred to as ‘zones’.
Discrimination between the protected line sections is achieved by having an instantaneous setting covering (or ‘reaching’) around 80% of the impedance of the first line section (zone 1). This precludes the possibility of overreaching into the second line section, thereby ensuring discrimination with the instantaneous protection on the second line section. A second time-delayed element covering around 120% of the impedance of the first line section (zone 2) ensures complete coverage for the first line section. Remote back-up protection for the second line section will typically be provided by a third element (zone 3) set to cover 120% of the impedance to the remote end of the second line section.
Instantaneous overcurrent and earth fault protection – most commonly used to provide protection of HV overhead line circuits against transient faults. The ‘instantaneous’ operation helps ensure the minimum arcing damage to overhead line conductors (e.g. due to clashing initiated by a falling tree branch) and hence the best possible chance of a successful auto-reclose operation.
Instantaneous protection is commonly used in conjunction with inverse definite minimum time (IDMT) protection on auto-reclose schemes such that after one or more unsuccessful reclosure attempts, the INST protection will be disabled leaving the IDMT protection to clear the fault if necessary (i.e. if the fault is not then cleared by downstream protection – e.g. by a downstream in-line auto-reclose circuit breaker or auto sectionaliser).
High set instantaneous overcurrent protection – is a variant commonly used on 33 kV overhead line transformer feeders. The ‘high’ setting ensures that the protection will not operate for faults beyond the 11 kV winding of the downstream primary 33/11 kV transformers.
IDMT protection – the most common form of protection used for overcurrent and earth fault protection of high voltage distribution circuits. The protection characteristic is such that its speed of operation is related to the magnitude of the fault current it detects, but with a ‘definite minimum’ time which determines the maximum speed of operation irrespective of the fault current magnitude. Both the operating current (e.g. ‘plug setting’ 150%) and time delay (e.g. ‘time multiplier setting’ 0.15) can be set to allow the ideal protection characteristic to be set in order to ensure both reliable detection of fault current and adequate discrimination. Relays with very inverse and extremely inverse characteristic settings are also available.
Standby earth fault protection – a type of IDMT protection applied typically to 11 kV switchboard incoming transformer 11 kV circuit breakers (and/or sometimes bus section circuit breakers). Its role is to provide back-up to the IDMT protection on individual 11 kV feeder circuit breakers in the event that the feeder protection fails to operate (or if the circuit breaker fails to trip). It follows that the SBEF protection settings must be set to discriminate with the feeder protection settings. SBEF is also the primary earth fault protection for the substation 11 kV busbars.
Substations will generally be equipped with two stages of SBEF protection – SBEF1 as described above and SBEF2. SBEF2 provides back up to SBEF1 and in addition to operating the incoming transformer circuit breaker(s) will also initiate an intertrip signal or the operation of the relevant local fault throwing switch (or both FTSs if the 11 kV bus section circuit breaker(s) have no protection). SBEF2 provides the means to clear an 11 kV feeder or busbar fault in the event of a flat tripping battery, or other fault affecting the tripping circuits, at the primary substation.
Directional overcurrent protection – as the name implies is a form of (IDMT) protection which is constrained to operate for power flows in one direction only. Its most common use is at 33/11 kV primary substations served by transformer feeders where its purpose is to clear a fault on the incoming 33 kV transformer feeder when the two (or more) transformer feeders are operating in parallel to supply the 11 kV switchboard.
The DOC relay will operate the incoming 11 kV transformer feeder circuit breaker on the faulted circuit in the event of fault current passing from the healthy 33 kV transformer feeder(s) to the faulted 33 kV transformer feeder via the 11 kV busbars (it will not operate for fault currents in the opposite direction and hence has no need to discriminate with 11 kV feeder circuit IDMT OC protection for 11 kV feeder faults). The directional characteristic is provided by means of a constraining torque acting on the relay derived from the 11 kV switchboard voltage transformers.
Because the path for the fault current is from the adjacent transformer feeder(s) through the 11 kV busbars, it follows that the fault current will be relatively small due to the impedance of the transformers through which it passes. A consequence of this is that the protection setting will need to be relatively low to ensure reliable operation, and care is therefore needed to ensure that any normal reverse power flows from the 11 kV busbars to the 33 kV system (e.g. due to 11 kV connected exporting DG) does not trigger operation. The most onerous condition in that respect is when DG is exporting when only one 33 kV transformer feeder is energised.
Neutral voltage displacement (NVD) protection – is essentially a means of detecting a line-ground fault on an unearthed system – i.e. where a fault has been cleared at source thereby breaking the path between a grounded conductor and the star point of the source transformer, and therefore the system earth.
The need to detect the condition arises where the grounded conductor remains energised through a separate source such as (possibly) a generator or (typically) an adjacent transformer feeder. For example, an earth fault on an overhead 33 kV transformer feeder would be cleared by the IDMT earth fault protection on the 33 kV source circuit breaker (generally following an unsuccessful auto-reclose attempt) but the line would remain energised via the adjacent 33 kV transformer feeder and the 11 kV busbars (note: the fault would not be cleared by DOC protection – see above – unless at least two conductors were grounded).
The protection detects the fault by means of either a VT or more commonly by NVD cones (capacitors) connected onto the incoming 33 kV conductors at the primary substation. No earth fault current will be flowing (other than that due to line shunt capacitance) so protection clearance times are normally set to several seconds to allow ample time for a transient fault to clear.
NVD protection is sometimes specified as part of the protection suite for a distributed generator as a back up to loss of mains protection. This is because when operating in parallel, the star point of the generator will have been disconnected from earth (so as not to create a multiple earthed system) and so a grounded conductor on the distribution system would not be detected by the generator earth fault protection.
Restricted earth fault protection – is a specific form of transformer unit protection whereby three of the four current transformers are normally positioned on the 11 kV tails of a 33/11 kV transformer at the point of connection with the substation 11 kV switchboard incoming transformer circuit breaker, with the fourth CT positioned between the (typically delta-star) transformer 11 kV winding star point and the 11 kV system earth (normally effected through a neutral earth resistor or reactor). The positioning of the CTs defines the protection zone. REF is a form of differential (or balanced) protection which provides ‘instantaneous’ protection for the bulk of the transformer 11 kV winding 1 and the transformer tails.
Buchholz protection – is a common form of primary protection for supergrid, grid and primary transformer internal faults. The relay is normally housed in a chamber positioned in the return oil feed from the top of the transformer tank to the conservator.
The Buchholz relay has two modes of operation. A gas alarm will be triggered if sufficient gas accumulates (suggesting an emerging fault condition in the transformer tank) or if the oil level drops below a certain point due perhaps to a leak. It enables early diagnostic assessments to be made (e.g. dissolved gas analysis) and if necessary the transformer to be taken out of service before a potentially catastrophic fault occurs.
A Buchholz surge relay will operate in the event of a short-circuit occurring within the transformer tank, which will initiate a surge of gas and oil, passing through the oil return pipe and the Buchholz chamber, out to an explosion vent. The relay will operate instantaneously to trip the local switchboard transformer incoming circuit breaker, and trigger the substation tripping relay to initiate either an intertrip signal or a FTS operation to trip the remote source circuit breaker. It will be evident from the diagram adjacent that, in the event of the oil level falling sufficiently to operate the lower mercury switch, this will also trigger the substation tripping relay (Fig 1).
Fig 1: Buchholz relay
Winding temperature protection – comprises both an alarm and trip function. In the event that the oil temperature inside the transformer tank approaches a critical temperature, an alarm is initiated warning the control centre staff that load may need to be reduced or shifted. Should the temperature continue to rise to a critical level then a trip relay will operate the local circuit breaker to offload the transformer.
Overall protection schemes
Having taken a look at the most common forms of protection, this section looks at typical overall protection schemes. Note that network companies will have their own standard protection schemes which may differ in detail to the examples shown.
Fig 2 is a simplified block diagram of a typical overall protection scheme for a 33/11 kV primary substation served by two transformer feeders (in this case one overhead and one underground). Note that some components have been omitted or simplified to avoid cluttering the diagram, for example ‘intertrip send’ and ‘intertrip receive’ relays have been omitted as have various alarms such as Buchholz gas, tapchanger out-of-step, winding temperature, etc.
Fig 2: Simplified protection block diagram for a primary substation served by two transformer feeders
Circuit unit protection
Whilst, as described in the previous section, most 11 kV feeders are protected through IDMT OC and EF protection (along with INST OC and EF2 in the case of overhead line feeders protected by auto-reclose circuit breakers) there are circumstances where a higher level of security than required by ER P2 (see parts 1 and 2 of the guide) is needed by customers due to the sensitive nature of their business.
Fig 3 depicts a unit protected ring. In the event of a fault anywhere on the ring, the circuit breakers at each end of the faulted section of cable at the two adjacent substations will trip – isolating the fault at both ends without any interruption of supply to customers served by the three substations forming the ring. The means of detecting a fault is by comparing the current entering each section of cable at one substation with the current exiting that same cable at the next substation.
Fig 3: Solkor unit-protected ring
It will be evident from Fig 3 that under healthy conditions, for any section of cable, the current flow will be in the same direction and the two measurements will be equal. In the event of a cable fault, the currents entering the faulted section of cable at each end of the cable will be in opposite directions.
For all other sections of cable, although the current flowing under fault conditions will be very much higher, the direction and magnitude of the current flow at each end of the healthy section of cable will remain the same.
The means of comparing current flow and direction is by use of current transformers fitted to the circuit breakers connected at each end of each cable section, and through pilot cables connected between each substation.
Under normal healthy conditions, the CT secondary currents entering each end of pilot cable section will balance. However, in the event of a fault, the currents entering the pilot cables from each end of the faulted cable section will not balance. This imbalance is detected by a relay at each circuit breaker initiating a circuit breaker trip at each end of the cable section.
Unit protected rings are costly to install due to the need for pilot cables and circuit breakers (as opposed to simple line switches) and so are generally justified only under circumstances where continuity of supply is critical.
Loss of mains protection
Loss of mains protection has become a common feature of distribution systems, particularly 11 kV and LV, due to the proliferation of DG and microgeneration. The requirement for LOM protection is described in Engineering Recommendations G59/3 and G83/2, and its provision (by the customer) is a requirement of the Distribution Code. Common forms of LOM protection are rate of change of frequency and vector shift. The purpose of LOM protection is to ensure that in the event of a fault affecting the distribution network to which the generator is connected, the generator will disconnect from the network.
Supplementary protection requirements include over voltage and over frequency, and under voltage and under frequency. The recommended settings are prescribed under G59/3 and G83/2 and compliance with these settings is essential to ensuring that the generation will reliably disconnect in the event of loss of a ‘loss of mains’ condition, whilst minimising the risk of disconnection in the event of a voltage dip or a sudden fall in system frequency.
Maximising the so-called ‘ride-through’ capability of DG in the event of transmission system faults or frequency excursions is becoming increasingly important due to reducing system ‘strength’ (fault level and inertia) arising from the ongoing displacement of large synchronous generation by converter connected solar photovoltaic generation, doubly-fed induction generators associated with onshore wind farms, and dc connected offshore wind farms. Specific fault ride-through requirements are included in the EU Network Codes now being introduced into Great Britain for all generators of 1 MW or greater size.
Voltage dependent and voltage controlled overcurrent protection
Overcurrent protection is a further important component of an overall generator protection scheme designed to trip the generator in the event of a line–line fault. However, an initial high fault current will cause the generator voltage to drop, so decaying the fault current potentially below the relay setting. To overcome this, voltage dependent relays incorporate a bias which progressively decreases the overcurrent setting if there is a voltage drop.
A variation on the voltage dependent relay is the voltage controlled relay which has both an under voltage setting and an overcurrent setting which is less than the rated current of the generator. The relay will operate only if the under voltage and overcurrent elements operate at the same time – i.e. in the event of a fault.
HV feeder protection
Fig 4 represents a fairly traditional arrangement for a rural 11 kV feeder protected at source by a primary substation circuit breaker. The first section out of the primary is underground cable serving a local underground distributor but with a short overhead line spur protected by drop-out expulsion fuses. The source circuit breaker protection will discriminate with these fuses so that in the event of a fault on the spur the fuse would clear the fault and the main circuit would remain energised.
Fig 4: Traditional arrangement for a rural HV system
Further along the feeder the circuit changes from underground to overhead construction. At this point an auto-reclosing circuit breaker is installed. Again the source circuit breaker protection will discriminate with the auto-reclosing circuit breaker (ARCB) so that transient or sustained faults beyond the ARCB will be cleared by the ARCB leaving the upstream main circuit energised.
In the event of a fault, the ARCB will follow a sequence of auto-reclose attempts to restore the circuit. If the fault is transient, then all supplies will be restored when the ARCB recloses. If the fault is sustained, and on the main line, the ARCB will lock out. If, however, the sustained fault is on the spur protected by the automatic sectionaliser, then the ARCB and auto sectionaliser will co-ordinate – i.e. the auto sectionaliser3will isolate the spur and the ARCB will then successfully reclose energising the remainder of the overhead line circuit.
Fig 5 is a typical (albeit simplified) arrangement for an urban or suburban electricity system supplied by underground cables. Distribution substations along the feeder are predominantly based on ring main units (RMUs) but with a number of tees. The feeder is protected by a source circuit breaker with IDMT OC & EF protection and has a normal open point (NOP) which it shares with an adjacent feeder to form an ‘open’ ring.
Fig 5: Traditional arrangement for an urban HV system
It will be apparent from Fig 5 that in the event of a fault on any part of the open ring system, the faulted section can be isolated by use of the RMU ring switches, and supplies subsequently restored by reclosing the tripped circuit breaker at the primary substation and the NOP.
In the event that the fault occurs on a teed section of cable, then the supplies served by the affected substation can often be restored from an adjacent substation through LV backfeeds (i.e. inserting links in link boxes). In some cases, where the LV backfeed is insufficient or non-existent, a mobile generation unit might be deployed.
In this way, the design permits compliance with the ER P2 capability requirement for a class B demand group to restore supplies within 3 hours (i.e. through manual switching). Substation capacity and demand on any teed section will normally be limited to 1 MW (i.e. class A demand group) in order not to infringe the ER P2 3-hour restoration requirement. An exception might be where the tee supplies a single industrial or commercial customer who has elected to accept the higher supply security risk in favour of a lower connection charge.
It follows that attention to overall network loading is necessary to ensure that the ‘alternative circuit’ is able to meet the full demand of the faulted circuit in the event of a fault occurring on the first section from the primary substation.
To aid identification of the faulted section of a feeder in the event of a circuit breaker trip, it is common practice to install fault passage indicators at each RMU. The fault passage indicator will indicate whether or not fault current has passed, and by process of elimination the faulted section can be determined.
Remote control and automation
Given the now strong regulatory incentive regarding quality of supply (see section on Quality of Supply in part 2 of the guide) DNOs are now extensively applying remote control and automation to their 11 kV networks (since 11 kV faults account for the great majority of customer interruptions and ‘minutes lost’ (CIs and CMLs). RMUs equipped with remotely controlled ring switches are now common, and these will often now form part of an automation scheme.
Switching by remote control will restore supplies far more quickly than by manual switching, hence reducing the CML impact of the fault. However, an automation sequence which completes within 3 min of the fault occurring will also reduce the CI impact as any interruption of less than 3 min is not counted as a CI under the regulatory quality of supply incentive.
Fig 6 demonstrates a typical automated supply restoration sequence for an urban network.
Fig 6: Urban network automation sequence
Remote control and automation is not unique to urban systems; it is now commonplace to apply similar schemes to rural systems (such as depicted in Fig 4) using remotely controlled auto-reclosers and sectionalisers. Such schemes will combine the benefits of automatic restoration for transient faults with the ability to rapidly isolate the faulted line section (or spur) in the event of a sustained fault. Automation of the switching sequence will enable restoration of supplies to all but the isolated section within 3 min.
- Albeit a very sensitive form of protection, winding faults very close to the star point might not generate sufficient fault current for the protection to operate
- Some systems do not use INST EF protection but may include sensitive earth fault protection to detect high impedance line-earth faults
- There are variations in the precise way that different auto sectionaliser technologies work