Based upon the previous chapter discussions and the observations emanating from the case studies presented, conclusions are drawn. Ultimately, there would be some suggestions regarding the future work to extend this research work further and, as a result, make it more refined.
9.1 Conclusions
The current power industry environment entails that countries around the globe commit or set them in the process of introducing more competition into their power industry regimes.
Upon moving from monopolies to competitive electricity markets, it was deemed imperative to ensure open access to transmission and distribution networks in an effort to promote effective competition in the electricity supply sector. In that regard, the charges for the use of network asset set by network companies are the central element in providing efficient economic signals for guiding the siting and sizing of oncoming demands and generators, and incentivizing the efficient use of these network assets. These network assets refer to lines, transformers, VAr compensation assets, protection assets, e.t.c.
In the UK and the rest of the world, there are many such economic use of network asset pricing approaches (network charging methodologies), for transmission and distribution networks, reflecting investment cost incurred in the network circuits and transformers to support real and reactive power flow, as mentioned in chapters 3 and 4. However, to reflect the investment cost for maintaining network voltages in network charges has received very little attention in network charges. Currently, power factor penalty approach is used to recover costs of operation, mainly generator fuel related cost, and does not provide for those of network VAr compensation assets for maintaining network voltages. This power factor penalty approach has been criticized by many researchers as it is not based on economic principle and, moreover, it is regarded as inconsistent and inadequate.
Furthermore, with the increasing penetration of embedded generators (EGs) in the UK, there is a growing need to formulate pricing methodologies based on economic principle to ensure that network assets are effectively and efficiently utilised [6]. Since the existing approach does not provide correct economic efficient price signals, EGs may potentially locate at sites which may result in considerable network investment. Therefore, it is against this background that, this research work is directed to address this issue of recovering network costs associated with maintaining the network voltages. Therefore, Long-Run Incremental Cost (LRIC)-voltage network charging principle was proposed and it is one of the major contributions in this research work.
LRIC-v network charging principle is based on the use of spare nodal voltage capacity or headroom of an existing network voltages to provide time to invest in VAr compensation devices. The VAr compensation devices, in question, are particularly SVCs given the issue of encouraging EGs penetration into the network. EGs mainly are wind turbines and they introduce voltage variations into the network and, therefore, for smooth system operation fast acting SVCs would be ideal to be introduced [6, 22, 23, 24]. The LRIC-v network charging principle was at first proposed to price for the future network VAr compensation assets and extended until it became comprehensive to be able to price for the existing network VAr compensation assets, among others. The evolution of this proposed novel pricing methodology over the course of the research is as follows:
1. LRIC-Voltage Network Pricing Reflecting Future VAr Compensation Assets
Firstly, this approach was proposed utilizing network voltage variations to reflect the future network VAr compensation asset prices resulting from nodal withdrawals/injections [63]. The most attraction of this approach is that for the first time a method to evaluate long-run incremental cost is proposed based on the spare nodal voltage capacity of an existing network. The resulting network voltage charging model is able to provide locational forward-looking economic signals, reflecting the extent of the impact to busbar voltages by a connected party, i.e.
whether they accelerate or delay the need for future network compensation devices. In the event of a network user accelerating the requirement for future network compensation assets, that user shall be penalized for that action. Otherwise, if the network user delays the requirement for future network compensation asset reinforcement(s), such a user shall be rewarded for that action. In addition, this approach considers both real and reactive power nodal withdrawals/injections. These economic messages will, in turn, influence generation/demand in order to minimize the cost of future investment in VAr compensation.
2. LRIC-Voltage Charges On The System With Different R/X Ratios
A fundamental study was performed to analyses the trend of LRIC-voltage network charges on different types of networks [77], providing insights into how charges would change with different network circuit resistance/reactance (R/X) ratios. The results showed that when the network circuit Xs are at least ten times more than their R counterparts, only MVAr nodal perturbations should be considered. When the network circuit Rs, on the other hand, are at least ten folds more than their Xs counterparts, then, only MW nodal perturbations should be considered. Finally, when the network circuit Xs and their corresponding Rs are comparable, both MVAr and MW nodal perturbations should be considered. In a nutshell, this meant that for transmission networks, only MVAr nodal perturbations should be considered while for distribution networks both MW and MVAr nodal perturbations should be considered.
3. How the LRIC-Voltage Network Charges Varies With Different Demand Load Growth
Another fundamental study was carried-out to analyse the trend of LRIC-voltage network charges on different demand load growths, providing insights into how charges would change given different demand load growths. The results showed that the LRIC-v network charges given different load growth rates are a function of the system nodal voltage loading levels. This means that, there is a nodal bus voltage loading threshold above which if most buses are loaded at in a power system, the least load growth rate would have more charges. On the other hand, below the aforementioned nodal voltage loading threshold and if most power system buses are involved, then the larger load growth rate would generate more charges.
4. LRIC-Voltage Network Pricing To Support Network Voltages Under N-1 Contingencies
The LRIC-voltage network pricing previously proposed was extended to consider n-1 contingencies, as it is a requirement that these types of contingencies should be considered to ensure acceptable network security and reliability. The investment cost-related pricing (ICRP) charging model [72] used in the UK, for recovering investment costs of network circuits and transformers, does not consider the network security requirement in their pricing model, but, it relies on post-processing through a full-contingency analysis to give an average security factor of 1.86 for all concerned network assets. On other works related to investment costs of circuits and transformers, authors of [65] demonstrated a simplistic approach to network security based on the assumption that reinforcement is required when a branch reaches 50% utilization. Authors of [66]-[73] considered the n-1 contingency analysis into their charging principles and all of these were for pricing of network circuits and transformers. In this regard, it was imperative to factor n-1 contingencies into the LRIC-v network charges. The results showed that the respective charges follow the same pattern as the original approach, but are increased since the nodal busbar voltage margins are reduced to accommodate n-1 contingencies. This, in turn, provides correct economic price signals since the network operators are required to operate their networks physically to accommodate n-1 contingencies.
5. LRIC-Voltage Network Pricing For Existing Network SVCs
The LRIC-voltage network pricing approach, for pricing future network SVCs, was extended to cover for pricing the existing network SVCs, where SVCs were installed in the network. The earlier proposed charging principle failed to price for the use of existing network SVCs since where these existing devices are sited the corresponding bus voltages do not vary. The reason why the bus voltage does not vary is because the existing network SVCs would be preset to keep the corresponding nodal voltages at constant values and, therefore, since the proposed LRIC-voltage charging principle depicted its strength upon the network voltage variations, this approach could not be applied in this regard. It is true that the reactive power loading of the SVC varies in keeping the bus voltage, where this device is sited, at a preset value given variations in power system demand and generation loading conditions. In this regard, this new extension involved the mapping of the SVCs‟
reactive power loading limits to the nodal voltage limits at the buses where these SVCs‟
are sited. For example, if the SVC reactive power lower and upper limits are QminandQmax, respectively and, on the other hand, if the nodal voltage lower and upper limits, where this SVC is sited, are V andL V , respectively. To that end, H Qminwas mapped to V while L Qmaxwas mapped toV . With this limit mapping exercise completed, then, it was possible H to price for the use of the existing network SVCs.
6. Improved Implementation For LRIC-Voltage Network Charges
This entails an improved version of the LRIC-v network charges. This improvement emanates from the premise that the voltage change at a node and its corresponding power (MVA) change are related to each other by the P-V curve. As such, the best approximation of this relationship was used to execute the nodal voltage degradation rates resulting from
the load growth rate and, finally, the improved LRIC-v network charges were sought. In the earlier proposed basic LRIC-v network charging approach, the respective nodal voltage degradation rates were computed by noting the nodal voltages before and after annual network load growth and fixed these nodal voltage degradation rates to signify the constant rates at which the voltage would be varied from the current voltage levels down to the respective nodal voltage limits. This improved version resulted in processing the charges based on nodal voltage degradation rates calculated to reflect the correct physical relationship of the nodal voltages consequent to the system load growth rate.
Overall, LRIC-v network charging principle is intended to allocate the network VAr asset costs based on the usage of the network nodal voltage capacities from study nodes. It achieves this by evaluating the future network VAr compensation asset investment requirements and fairly allocates the future network VAr compensation costs to users. The strong points of the LRIC-v network charging principle incorporate the ability to reflect the forward-looking charges, to distinguish the charges at different locations, to consider both real and reactive powers, and to derive charges for both demand and generation users.
Finally, the LRIC-v network charging approach appeal to all stakeholders, namely, network owners, and network demand and generation users. From the perspective of network owners, it can reduce the future VAr compensation asset costs while at the same time improving the overall network voltage profile. From the perspective of demand and generation users, it can provide the lowest possible use of system charges. Moreover, the reduced electricity prices for the end consumers.
9.2 Future Work
In this section, some fundamental extensions to the work done in this thesis shall be recommended to ensure that the developed work can be more suitable in real world network charging methodologies. Those shall be briefly expressed, each in turn, below.
1. The use of electrical distance concept for effective voltage control
Given the densely meshed nature of the power networks nowadays and the recent voltage problems in a number of power network establishments, it is apparent that to solve the voltage problem, the concerned network ought to be divided into voltage control zones as explained in Appendix A. This approach entails the use of electrical distance concept to identify effective distinct network voltage control zones to independently master the voltage control in each zone as need arises. What set apart this approach from the rest, is that, it deals with the physical structure of the network as opposed to the trial and error method entailed in the conventional approaches. In Appendix A, a rigorous literature review was undertaken and, as a result, the use of electrical distance concept complimented by GA, as the decision-making tool, was found to be the most attractive integrated framework given the aforementioned current power industry conditions, for the siting of
SVCs to achieve an overall effective network voltage control. Therefore, this approach should be ventured into and, thereafter, all possible contingencies should be considered to ensure the proper sizing of the SVCs, which would constitute the whole exercise of RPP process.
2. Interaction between thermal and voltage network charges
The upgrading of the network consequent to either power withdrawal / injection can be dealt with by either constructing a new line / installing a transformer (resulting from LRIC-thermal network charges) or installing a reactive power compensation device (resulting from LRIC-voltage network charges). It should be noted that this exercise can also, at times, be achieved by either installing a higher rating of any of the aforementioned devices.
Both of these approaches, in turn, can improve the overall network circuit power carrying capacity and its voltage profile. In this regard, a comprehensive spot-on interaction between these two approaches has to be established to strike a balance between cost effectiveness and the most effective way to ensure network security and reliability as per statutory requirement. Therefore, it is against this background that, a rigorous study has to be instituted to establish this aforementioned interaction between thermal and voltage network charges.
3. Improving more the LRIC-voltage network charges
The load growth was assumed to be 1.6% in determining the LRIC-V charges. This factor compromise the accuracy of the charges as in reality different buses have different load growths. In future if the respective load at every load bus can be sought then the charges can be more accurate. Ultimately, this would institute some accuracy in the LRIC-voltage network charges.
4. Proper selection of the slack bus
It is recommended that before the LRIC-voltage network charges can be evaluated for any given network, the correct location of the slack bus should be established to represent the physical reality of the concerned particular network relating to the same. LRIC-voltage network charges are different at every different slack bus position on any chosen network, therefore if the slack bus is wrongly placed, incorrect economic signals would be sought and those would not be reflective of the situation on the ground.
5. Factoring n-2 contingencies into the LRIC-voltage network charging approach In this research work the network security is assessed given n-1 circuit contingencies. This means that each circuit is outaged and replaced back in its intact state, each in turn, to ensure that the network can withstand this kind of situation. Specifically, the effective network nodal voltage limits are established given the worst possible n-1 circuit contingency to ensure that the proper network nodal voltage ranges are established in order to evaluate the correct LRIC-voltage network charges which reflect the aforementioned contingency scenario. This emanates from the fact that, in reality, the network is expected
to be able to withstand this kind of contingency scenario to ensure that voltage instability is not experienced. Given the aforementioned, this contingency situation should be reflected in the LRIC-voltage network charges. On the other hand, P2/6 involves the assessment of the system security contribution from distributed generation given n-1 and n-2 circuit outages. The latter outage situation involves removing any two network circuits at a time and replacing them back, in turn, while assessing the concerned system security contribution from the distributed generation. Since in the UK, the government made commitment to encourage the distributed generations to connect onto the system, then it would suitable to factor into the n-2 circuit contingency situation in the LRIC-voltage network charges which is recommended for future work.
Finally, it is hoped that this thesis would raise attention in the area covered and, ultimately, stimulate further research in the aforementioned area.