AMAZON multi-meters discounts AMAZON oscilloscope discounts
[cont. from part 1]
4. SYSTEM PARAMETERS
4.1 Distribution Feeder Arrangements
Typical distribution feeder arrangements are shown schematically in Fig. 9 and described below:
_ Simple ring arrangements as used from primary or distribution switched substations offering a high level of supply security. Under fault conditions the faulty part of the ring may be isolated either manually or automatically and power delivered to the load via the healthy part of the circuit.
_ Interconnected or three legged ring arrangements as used from primary or distribution-switched substations. Used where the use of two simple rings is not possible geographically or where the load over the period under consideration does not warrant the security of supply offered by two separate ring-feeder arrangements.
_ Radial tee off a simple ring as used where the load is small and where it is not economic or practically feasible to include a ring feeder arrangement.
_ Radial feeders from primary or distribution switched substations. May be used either where it can be forecast that the future load growth and associated extensions will warrant the eventual formation of a ring or where it is geographically impossible to provide a supply from a radial tee. An additional use of a radial feeder is in conjunction with an auto-recloser/ sectionalizer scheme.
_ Express feeder as used to establish a distribution-switching substation as a sub-distribution point in the network. In such cases the distribution sub station should employ a bus-section switch and have two or more incoming supply sources from the same primary substation such that a 'firm' supply is established.
_ Interconnected distributor as used for essential (or very important person, VIP) services. In such cases further security of supply may be obtained by infeeds to the primary substation from different parts of the grid system.
_ Interconnector as used to allow a partial alternative source to a distribution substation. In the same way as for the interconnected distributor, increased security of supply may be obtained by infeeds to the primary substation from different parts of the grid system.
_ Subring as used to supply areas where a simple ring may be geographically hampered. By the careful positioning of isolators it is possible to isolate any fault within a small section of the circuit thereby enabling restoration of the supply to the remainder of the customers.
The designed rating of new circuits within any of these configurations must take account of:
_ The maximum load to be expected under normal network conditions within the period of planning, based on the load forecast.
_ The maximum loads to be expected under emergency conditions, given the chosen network reliability criteria.
_ The maximum and minimum voltage conditions allowable at all points on the network.
_ The possible need to deal with abnormal load flows resulting from distributed generation within the network.
_ Medium- and long-term network developments that may affect the circuits concerned.
_ The need to balance the financial penalties of investing in advance of requirement (see Section 22) with the social or environmental difficulties of repeating work in the same area.
4.2 Voltage Drop Calculations
The vector diagram in Fig. 8a applies to a three phase system where ES and ER are the phase to neutral sending and receiving voltages, respectively.
I is the line current of the three phase load. R and X are the line resistance and reactance with cos Ø being the system load power factor.
====
FIG. 8 Distribution planning load forecast (The graph postulates
a forecast prepared in 1994 with a forward projection to 2010).
====
FIG. 8a Vector diagram.
FIG. 8b Distribution factor.
FIG. 9 Typical distribution feeder types.
Useful values of power factor for distribution systems that are often encountered are given in Table 3.
4.3 Positive Sequence Resistance
The resistance of cables is dependant not only upon the physical make-up and conductor materials but also the operating temperature. Reference should be made to Sections 12 (Cables) and 18 ((Overhead Line Conductor and Technical Specifications). Based upon a 20 C everyday working tempera ture the AC resistance, RAC ohms becomes:
4.4 Inductive Reactance
4.5 Economic Loading of Distribution Feeders and Transformers
Section 22 (Project Management) explains how investment decisions may be appraised using simple financial and economic analysis tools. The effect of inflation and the high cost of borrowing money to finance a project means that great importance must be paid to matching the initial capital expenditure for the distribution equipment against the future revenue stream resulting from the sales of electrical energy over the lifetime of the project. Therefore equipment must not be oversized or over specified. At the same time allowance must be made for the equipment to be capable of dealing with the factors listed in Section 4.1. Hence the importance of distribution planning and data collection to allow the utilization factor, load factor and other key parameters as described in Section .2 to be evaluated.
It is most important to be aware that discounted cash flow techniques in themselves do not give a 'correct' investment and distribution design answer.
Such techniques should be used for comparative purposes with the goal of maximizing the returns in line with the electricity supply utility's performance measures and also minimizing depreciation, technical load and no load losses, non-technical losses, maintenance costs, taxes, etc. Sensitivity analysis allows the supply utility to determine the minimum revenue requirements to support the proposed expansion project.
4.5.1 Annual Feeder Costs
The total annual cost of a feeder per phase, per unit length, CFeeder, may be evaluated from an expression taking the form:
CFeeder 5a1bL2 Feeder 23:6 where a5function (annual fixed costs per unit length per phase) b5function (annual cost of load losses in d per load2 per phase per unit length) no-load losses may be assumed negligible LFeeder 5total three phase feeder load (kVA)
4.5.2 Annual Transformer Costs
Annual transformer costs, CTransformer, take a similar form to the annual feeder costs but also take no-load losses into account:
CTransformer 5d1eL2 Transformer 23:7 where d5annual fixed costs e5annual operating costs taking no-load losses into account LTransformer 5transformer load (kVA) 23.4.6 System Losses System losses may be categorized as:
1. No-load power losses (transformer magnetizing currents, etc., see Section 14).
2. Load power losses (I 2 R copper losses, eddy current losses, etc.).
3. Reactive losses (poor power factor, transformer losses, etc.).
4. Regulation losses (voltage drops).
5. Non-technical losses (illicit connections, poor tariff collection or metering).
The relative importance of these losses at different parts of the overall power system is illustrated below:
Part of Reticulation Overall Losses (%) Total Losses (%) Annual Capital Expenditure (%) Generation 10 5 9.5 63 Transmission 30 1.5 312 Distribution 60 3 625 In order to help simplify load loss calculations the sum of the loads connected to an approximately uniformly loaded radial feeder may be lumped together at a set point along the feeder length. Consider the radial feeder below:
5. SYSTEM RELIABILITY
5.1 Introduction
Historically system availability has been assured by the use of heavy duty equipment and where necessary by the provision 'firm' supplies. A 'firm' supply point in a network is one where an outage due to a fault or during maintenance on one part of the system will not prevent a supply being avail able at that point. Duplication of equipment and alternative feed arrangements allow the supply to be restored either after a manual switching interval or automatically by the use of suitable switchgear, protection and control.
With advances in modern equipment manufacturing and reliability, coupled with less frequent maintenance requirements, it is possible to avoid full duplication of equipment and still effect an acceptable availability of supply.
When applying such principles consideration must be given to consumer satisfaction and the level of supply availability that consumers will still find acceptable. In some countries (e.g. UK), minimum standards of reliability are established.
Tables 4 and 5 detail typical input and output data upon which reliability studies may be based. Table 6 is an incident list for a feeder serving 200 customers (1,000 kVA of connected load).
5.2 Reliability Functions
5.2.1 Introduction
This section defines some terminology used by the American EPRI organization in distribution system reliability analysis. Examples, based upon the statistics shown in Table 6, are given to illustrate the usage of such data.
The generalized formulae which describe each of the factors given below are included in order to allow the reader to program them into a desktop computer.
5.2.2 System Average Interruption Frequency Index (SAIFI)
This factor describes the historical interruptions performance of the system.
On average customers would expect to have between one and two interruptions during the year.
====
TABLE 4 Reliability Studies Data Collection
Parameter Input Data Recorded Outage date, time and duration Cause Feeder/other equipment (designation or type) Weather No. of customers affected Division and district Comments/remarks Component that failed (identification number) Substation Voltage level Isolation component Pole/manhole, etc.
Fuse/switchgear data Other location reference Overhead line underground, etc.
Action taken Multiple restoration Major electrical component Phasing Component that failed (data) Address of outage Amount of lost kW/kVA affected Manufacturer of failed component Type of damage Protective device failure details
====
TABLE 5 Reliability Studies Output Data
Parameter Output Data Outage listing Subdivision by cause Performance indices (total system subdivisions, summaries and breakdown) Feeder circuit trouble list Components that fail by cause Distribution of incidents, based on duration Voltage level Detailed location (feeder circuit by cause) Performance in time periods by cause Protective component by subdivision of system Detailed location/feeder circuit by protective device Weather by subdivision of system
====
5.2.3 System Average Interruption Duration Index (SAIDI)
This factor describes the average duration of the interruptions or outages on the system.
Note that for the second incident described in Table 6 service restoration require two steps. One hundred customers were without service for 2 hours and 50 customers were without service for 10 hours. Each customer was without service for 4.55 hours during the year.
5.2.4 Customer Average Interruption Duration Index (CAIDI)
This factor describes the average customer outage duration.
CAIDI Total customer-hours interrupted; Total customers interruptions
TABLE 6 Incident Listing for Feeder Serving Some 200 Customers (1,000 kVA Total Connected Load)
5.2.5 Average Service Availability Index (ASAI)
This factor describes how closely the customer demand was met based upon a normally anticipated full 8,760 hours of supply availability in the year.
5.3 Predictability Analysis
In order to have some knowledge about the reliability of a system component and when failures might occur, it is first necessary to collect historical data. In this way, and for the particular application, the following questions may be addressed:
_ How often does the system fail? (frequency)
_ How long does it take to restore the system after failure? (duration)
And, of course, the distribution planning engineering needs to appreciate how much the system reliability is improved by a given action on a cost/benefit basis in order to aid investment decisions. Similarly, in contracts where system reliability is a contract condition, it is essential for the contractor to know the level of investment needed in order to meet the requirement, and for the client to satisfy himself that the proposal offered will satisfy his needs.
The distribution planning engineer needs to understand when such a transformer may fail in similar service conditions. This in turn might imply answering the following types of questions:
_ How often will it be damaged (by a falling tree, vandalism, etc.)?
_ How often will power surges (lighting, etc.) occur that might cause it to fail?
_ Are there any special conditions concerning this component?
In essence the time to failure for a particular installation is a random vari able. However, practical precautions may be taken to increase component life since, as explained in Section 14, it is well known that insulation failure occurs more rapidly at higher operating (long overload period) temperatures.
From the data:
A distribution system may be reduced, from its source to the load, to a sin gle equivalent component with a composite failure rate (?, failures per hour or per year) and a restoration time (r, meantime to restore supply in hours).
From this example it is seen that loss of LV supply is, of course, dominated by LV busbar failure since failures in other parts of the system are off set by the system having parallel supply paths. When studying such cases it is also necessary to consider the probability of both scheduled and a limited forced outage occurring simultaneously.
Scheduled outages are when equipment is deliberately taken out of ser vice, for example during maintenance or testing operations. Forced outages are due to component failure or faults.
A range of proprietary 'System Reliability' software is available to enable computer analysis of more complex and extensive systems such as the total system illustrated in Fig. 9. Some software uses the calculation method illustrated above other packages use more sophisticated probability approaches. In a situation where reliability is a contract condition, care must be taken to ensure that the methodology used meets the required specification. Whatever the method, users must remember that the accuracy of such predictions depends very significantly on the accuracy of the data input, and that the use of equipment fault rates or fault restoration times obtained from another system or even from 'typical' statistics may be quite inappropriate in any one given power system.
6. DISTRIBUTED GENERATION
6.1 General
The introduction of small generation units, typically of less than 5 MW each, into a distribution system has a significant effect on network design. Such distributed generation (DG) also called 'embedded generation' may be an aspect of a smart grid (see Section 27), but often the generation is installed without any plan for, or long before realization of the essential control systems of a smart grid. Existing systems are normally wholly or mainly passive, with control of voltage conditions and power factor limited, and control of frequency provided by central generation.
As levels of DG penetration increase in a network, operational problems arise. Normal or peak directions of power flow may be altered, and in addition the system becomes significantly less passive. For example:
_ Wind generators are usually induction units, which export real power but absorb reactive power.
_ Photovoltaic (PV) units operate through inverters, introducing harmonics (see Section 24). Domestic units may be as small as 1 or 2 kW each, but they may be widely spread.
_ Commercial or industrial combined heat and power (CHP) sites may use synchronous generators, and their mode of operation and reactive power exchange with the network may vary depending upon tariff constraints or the particular needs of the load at the site concerned.
Reference provides comprehensive details of more types of distributed generation and their effects on the network.
The key to managing the connection of DG is the provision of full and early information to the network designer and operator, so that the matters covered in the following sections can be safely handled, employing the necessary protection, load flow, fault level and harmonic studies.
6.2 Protection Aspects
The connection of generation within an MV network will have an effect on the operation of protection; in general, the current in a fault on the network will increase as the generator(s) feed current into the fault as well as the nor mal grid source, but the fault current at the source circuit breaker may be decreased, depending on the fault location and on the relative impedances of parts of the network. Discrimination between cascaded protection and also between auto-reclosing units will be affected. It may therefore be necessary to revise network protection settings.
Another problem is that under some circumstances directional overcurrent protection may operate due to the power flow from the generator toward the normal grid infeed, and consequently, the protection approach for the net work may have to be reviewed.
Where a system fault causes the grid supply to disconnect, it is possible that the fault current from a small connected generator may not be sufficient to trip the generator, leaving a hazardous condition. A key feature of current regulation in both the UK (G59/2 and G83) and in the USA (IEEE 1547-2003) is that smaller generation units must have protection fitted that disconnects them if grid supplies are lost. The most common form of this protection is a Rate of Change of Frequency (ROCOF) relay, though under-voltage or over/under frequency relays may be used. Fitting this protection also reduces problems on feeders with auto-reclosing, as the genera tor will trip at the first 'dead' sequence of the recloser.
If grid failure disconnection is not applied, the resultant islanded network section needs to have the controls known as a 'smart grid' see Section 27.
This approach can even be applied to a low-voltage network as a 'microgrid' if the total installed DG is strong enough although this would not be in line with current US or UK practice.
6.3 Feeder Loading
Local generation affects power flows in a network. Generally, the demand on grid infeeds will reduce, and loadings of feeders outward from the grid infeed will similarly decrease. System losses will decrease. In some cases, this can defer reinforcement requirements (see Section 6.4), but as DG increases in a network, generation injection can exceed demand in the immediate locality, and net load flow is then toward the grid infeed rather than away from it. In the extreme case, this net generation power flow can be the basis of the peak loading on a circuit, and be the driver for network reinforcement. Similarly, if even a small level of DG is connected at the end of a tapered distribution feeder system, losses can increase rather than decrease.
These changes in current flow can also interact with line drop compensation (LDC) where this is used at distribution substations. Consider a distribution system with several feeders from a grid infeed, only one of which has DG connected. If the total generation is sufficient to significantly reduce the total demand on the infeed at time of peak load, the LDC (which is designed to control voltage at the supply point by a signal to the transformer tap changers that is derived from a measure of system loading) will not increase sending voltage to the level it would have done if there were no DG. But the other feeders will still be experiencing their normal peak load and voltage drop along them will be as before. Without adequate voltage compensation at the supply point, consumers at the end of these other feeders may experience an unacceptably low voltage.
It is apparent that different balances between generation and load on such a network can have different effects. For example, the generation may not be sufficient to significantly affect the LDC operation at the infeed point, so that at times of peak load, instead of the voltage on some feeders being too low, the voltage at the end of feeders boosted with DG may become excessive.
Another problem with tapchanging systems, which arises if the power flow from the generation is sufficient to reverse the power flow through the incoming supply transformers at certain times of the day, is that the diverters of some transformer tapchangers are not designed for switching during reverse power flow.
6.4 VAr Compensation
Depending upon the nature of the generation, VAr exchanges into or out of the distribution network may occur, and studies should be undertaken to determine whether compensation equipment is required to maintain voltage conditions or stability (see Section 1). Fixed or switched devices are only suitable solutions in the simplest of cases, but static VAr compensators (see Section 25, Section 25.3.3) or STATCOM units (see Section 25, Section 25.3.4) and the other options mentioned in Section 27, Section 27.4 offer very flexible alternatives, which can take account of the possible intermittent and variable effects of DG. As in all engineering, however, cost is a vital factor and a sophisticated but more expensive solution needs to be justified in terms of the size of the problem.
6.5 Power Quality
PV units operate through solid state inverters, as do some wind generators, and these inject harmonics into the supply network. Reciprocating generators (such as those fed on waste-tip methane) can sometimes cause flicker or voltage fluctuations and fixed speed wind turbines have been reported to cause cyclic variations in the generator output that can lead to flicker.
Connection and disconnection of generation in a weak distribution network can cause transient voltage variations. Solutions to these and similar power quality problems are discussed in Sections 24 and 25.
On the other hand, the higher fault levels resulting from adding DG can in some cases improve power quality by reducing the effect of disturbing loads.
FIG. 10a Basic 33/11 kV substation supplying 18 MVA load at peak.
FIG. 10b Addition of load and local generation providing 4MVA at
peak.
FIG. 10c Effect of distributed generation with 50% availability.
Note: Sketches are simplified to illustrate principle; losses and
reactive power balances are ignored.
6.6 System Reliability
As explained in Section 5, availability of supply is nowadays the subject of detailed analysis. The addition of one or two small generators in a distribution network cannot be considered to add significantly to network reliability or supply availability, not only because of its likely automatic disconnection under emergency conditions (see Section 6.2) but also because its individual unit reliability is likely to be unknown or very low.
However, as the penetration of DG increases, the position may be different.
Consider the simple network of Fig. 10a. Under normal conditions, the losses are minimized by loading each 33/11 kV transformer to less than its nominal 12 MVA rating, but under maintenance or emergency conditions, the whole system load can be taken on one transformer. The capacity of the network to meet its group demand, which can be formalized as:
_ the cyclic rating of the remaining supplying circuits, following loss of the most critical circuit;
_ plus the transfer capacity that can be made available from alternative sources;
_ plus the effective contribution of any connected generation is adequate simply on the basis of the cyclic rating of the remaining infeed circuit.
If there is no transfer capacity when load grows, as in Fig. 10b,itis necessary to add more transformer capacity. For standardization and load sharing it will probably be another 12/24 MVA transformer that is added.
This is expensive and may in turn bring problems of high 11 kV fault level that may need the third transformer switch to be normally open and automatic reclosing to be added. The existence of 4 MVA of local wind generation that can support the new loads under normal conditions cannot be considered for emergencies because it cannot be relied upon to be available at the critical time it is not considered to be an 'effective contribution'.
However, significant penetration of small DG, as illustrated schematically in Fig. 10c may allow an alternative approach. An availability analysis such as that shown within Section 5.3 or as provided by proprietary software taking account of the fact that there are several units each with an individual low availability, may well show that the overall effective contribution of the generation is adequate to meet the required standards of continuity of supply to the consumers on the network.
7. DRAWINGS AND MATERIALS TAKE OFF
Drawings of the distribution network are normally maintained on a computer system using digitized maps and 'layered' data as described in Section 12.
These maps show the routing of the distribution overhead lines, cables, sub stations and feeder pillars on the background of the normal street maps.
Each and every part making up the system is given a unique identifier which links into the planned maintenance regime adopted by the utility. Such an approach also allows for the collection of statistical data for predictability analysis as described above.
A utility or consultancy will have developed standard ways of meeting its power distribution requirements over the years. Standard drawings linked to a computer database for such items as pole-mounted transformer arrangements, cable terminations, etc. and all the associated fittings which make up such an assembly will be recorded in this way. Maintenance or erection of new facilities then becomes a matter of planning the work (often using Program Evaluation Review Techniques (PERT) with bar charts and critical path analysis), drawing out the required parts form the stores and programming the work into the overall plans.
The Westinghouse 'CADPAD' program is an advanced computer-aided planning system covering all the main areas required. Such a program allows for long-term planning of new feeders and may be used to maintain a distribution planning database giving feeder connections, capabilities, lengths, substation locations, etc. The program may be used to assist with the determination of alternative feeder arrangements taking into account switching, reinforcements and new feeders for minimum cost within existing trans former capacity, minimum cable lengths, minimum voltage drop, etc. It produces a substation summary with loadings, and maximum voltage drop, etc.
It produces a substation area so as to highlight possible future reinforcement requirements.
Load flow and fault level analysis is also part of the package with auxiliary programs covering regulation (reactive compensation), reliability analysis and protection co-ordination. Further the program may be used to hold system constraints and a log of equipment data for stores and ordering purposes. The program may be coupled with an interactive digitizing system to allow distribution planning drawing to link into the overall design.