Discussion on the Impacts of Maintenance

Maintenance is of considerable impact on the deterioration/recovery behavior of a component in terms of condition improvements. Consequently, one may be interested in investigating the effects of maintenance on the CB deterioration/recovery status implementing the proposed methodology. In this regard, the following considerations are made [163]-[169], [181]:

 CB, as a component, is assumed to have four deterioration/recovery states, introduced earlier, where in the vulnerable state, the CB will still work properly.

The objective is to keep the CB at least working in vulnerable state, and look for timely maintenance.

 In the troubled state, the CB could still work but on the edge of failure. In the failed state, the CB may or may not work as expected; the open/close operation is not reliable at all and a large breaker operation delay may exist.

 There are three types of maintenance states assigned: minor for vulnerable condition, major for troubled condition, as well as failure repair (replacement) for failed condition. Take into account that these are all among the preventive maintenance considerations and not the corrective maintenance actions, which are commonly done on the CB after it fails.

 Maintenance should not turn the CB into a worse state, i.e., maintenance activities are assumed to be judiciously applied with no drawbacks. So, the states can be recovered/ improved via maintenance/repair.

133

 Minor maintenance will only bring the CB into the prior deterioration/recovery state; for instance, troubled state to vulnerable state, but will not lead to the healthy state directly from the troubled deterioration state. A major maintenance action can bring CB to healthy state.

 Minor and major repair actions have very small impact in turning the failed CB into a healthy state since a replacement has to be considered if aiming so.

Taking the above introduced assumptions into account and assuming the data presented in [166] and [169], the effect of different maintenance policies on the CB in different deterioration/recovery states can be quantified as demonstrated in two case studies in Table 12 and Table 13. It is obvious that the probability of successfully bringing the CB into a better working state following major maintenance is larger than following the minor maintenance [184], [185].

As noted earlier, maintenance attempts are devoted to keep the CB being at least in the vulnerable state (i.e., the sum of probability of the D1 and D2 larger than 85% after maintenance) before any action is taken. In the first case in Table 12, the analysis shows that the CB is very likely to be in a faulted situation. Even after a major maintenance, the total probability of D1 and D2 is 42.61% which implies that a repair is in an urgent need.

In the second case, the analysis results demonstrate that the CB is in its pretty good working state with a very small possibility of having damages. A minor maintenance could bring the CB into a state with a 90.5%, as the total probability of D1 and D2. However, a major maintenance, as a means of preventive action, which may cost more than ten times of the minor one, will only improve the overall reliability performance a bit better [166].

134

Table 12. Effects of Maintenance on the CB Deterioration/Recovery: Case I Deterioration/Recovery State Probability

D1 (%) D2 (%) D3 (%) F (%) Before maintenance 0.28 13.21 24.40 62.11 Minor maintenance 9.53 21.04 10.43 59.00 Major maintenance 35.99 6.62 7.41 49.98

Failure repair 99.00 0.91 0.08 0.01

Table 13. Effects of Maintenance on the CB Deterioration/Recovery: Case II Deterioration/Recovery State Probability

D1 (%) D2 (%) D3 (%) F (%)

Before maintenance 20 60 15 5

Minor maintenance 62.00 28.50 4.75 4.75 Major maintenance 88.00 7.85 2.00 2.15

Failure repair 99.20 0.78 0.02 0

Thus, one could make a conclusion that the proposed model will not only identify the cause of deterioration in a timely manner, but it could also give an optimized economic maintenance solution.

The following points are worthy to note regarding the observations made:

 The calculated failure probability of the CB (more than 50%) reflects that some parts of the CB under study responsible for the trip operation are not reliable enough and need maintenance. The failure probability index calculated here in real-time is different from the failure rate index (number of failures per year) commonly used in system reliability studies. Failure probability used here reflects

135

the CB condition at any given time. As a result, the two indices are different in unit and order of magnitude (failure rate is usually a very small value).

 Historical records are initially used to get an estimate of the CB reliability index as proposed in this Section. Such data may be acquired during time and with the operation of the CB by the monitoring devices. The probability distribution will then be updated as new data arrives. This gives a dynamic update of failure probability of CB in real time as time passes.

 The proposed approach considers the CB failures mainly related to the breaker timing parameters that can be recognized by the control circuit monitoring signals.

Some mechanical failures in the release and operating mechanisms as well as contact overheating, dielectric breakdowns, nozzle damages, incorrect assembly after maintenance, etc. may also affect the timing of the switching operation.

Moreover, sudden mechanical breakdowns, corrosion, leaks, and a wide variety of different wear and tear processes may be very difficult to detect. The research presented in this Section of the dissertation helps where the control circuit signals are the only or major source of monitoring data available in utilities, which is not rare in practice today. This also highlights the need for more comprehensive research on predictive asset maintenance management.

 Conducting the same procedure for every CB in a substation would lead to a reasonable differentiation of CBs maintenance scheduling and asset management practices. Making such a distinction between different CBs throughout a power system would definitely open new opportunities for the cost-effective asset

136

management decisions. It may be of interest to conduct a cost/benefit analysis to assess the economic aspect of putting the proposed approach in practice in the future. The gained economic benefits of the proposed monitoring scheme should outweigh the costs of device replacement and installation, fault detection, and assessment of CB condition monitoring over time.

5.5 Conclusions

The followings are some advantages of contributions elaborated in this Section:

 A quantitative approach to assess the reliability status of individual CBs and its subassemblies in real time is proposed.

 The proposed methodology uses the field monitoring signals from CB control circuit, which takes the advantage of increasing deployment of smart sensors and monitoring devices in the system.

 The presented approach allows a quantitative assessment of CB status leading to the classification into different deterioration/recovery states in real time.

 The real-time deterioration/recovery states differentiate the status of all the CBs in the system, which in general, is a helpful and reliable criterion as time elapses.

 Deterioration-based distinction helps in improving the system-wide maintenance scheduling and asset management practices to answer where, when, and how to perform the maintenance tasks on the system CBs.

137

 While only monitoring of CBs is considered in this research work, the same concept can be extended to other power system assemblies, e.g., transformers, insulators, etc. [186], [187].

 Since the most critical prerequisite for reliability and maintenance analyses is the past historical records, benefiting from a data gathering structure is an imperative.

This, indeed, is in line with the increasing trend in system-wide application of monitoring devices.

 This CB monitoring approach helps a more reliable and informed decision making for TLS implementation in practice as it unfolds the reliability of the associated CBs and, correspondingly, the availability of the transmission lines selected for optimal TLS implementations.

138

6 POWER SYSTEM TOPOLOGY CONTROL DECISION MAKING SUPPORT

TOOL FOR CORRECTIVE RECOVERY AND ENHANCED RESILIENCE^*

6.1 Introduction

Bulk electric transmission systems have been traditionally characterized with static assets and fixed configuration over time except in the cases of faults and forced outages.

Power system topology control, often called transmission switching (TLS), offers the system operators an opportunity to harness the flexibility of the transmission system topology by temporarily removing transmission lines out of the system. By changing the way how electricity flows through the system, TLS can be employed in emergency scenarios to alleviate violations, congestions, and overloading conditions. Such considerations, which are employed in the operational time frame, make it possible to have more efficient use of the existent network facilities.

Though being performed for decades on a very limited scale with rather focused aims, transmission switching has recently gained further importance with the increased penetration of renewable energy resources and the growing demand for more reliable operation of power systems. It has been shown that various operating conditions can be resolved through transmission switching; amongst, one can mention voltage violations and

* Part of this chapter is reprinted with permission from “Flexible Implementation of Power System Corrective Topology Control” by P. Dehghanian, Y. Wang, G. Gurrala, E. Moreno-Centeno, and M.

Kezunovic, Nov. 2015. Electric Power System Research, vol. 128, pp. 79-89, ©2015 Elsevier.

139

overloading conditions as a result of possible contingencies, network losses and congestion management, security enhancement, reliability improvements, and also cost reduction for economic benefits [50].

Many of the past research works on corrective TLS have suggested optimization procedures that provide one switching sequence and involves repetitive solution of transmission switching optimization until a valid switching plan satisfying the AC feasibility and stability is found. However, if the selected transmission lines are not switchable due to the associated circuit breaker (CB) failures, the switching process stops and the operator may need to re-run the optimization engine to obtain a new switching sequence. In practice, transmission line switching implementation involves several operational procedures and clearances at various levels of the utility organizations which are commonly time consuming. So, transmission system operators (TSOs) need to be provided with switching plans with high probability of success to minimize the involved time and labor. To the best of the author’s knowledge, none of the existing topology control algorithms address such practical considerations.

Trying to bridge the gap between the theoretical advancements in previous literature and the practical requirements that the operator will have to deal with, this Section proposes an implementation procedure for realizing TLS in practice. The main objective of this Section is to demonstrate how such technologies can empower the operator not only to obtain feasible switching actions, but also allow the operator to use his/her experience and personal judgment to decide which feasible set of actions to implement. The proposed framework, to be used in day to day operation planning

140

scenarios in response to probable contingencies, provides several switching options per contingency in a tree-like structure. At each level of the switching tree, the framework suggests the operator a set of AC feasible and stable TLS plans based on a selected optimization criterion such as maximum load shed recovery (LSR) or minimum generation cost, etc.

This Section also proposes a decision making support tool based on a CB reliability assessment technique using condition monitoring data. A mean benefit index is proposed to quantify the impacts of failed CBs associated with switching of a transmission line in any substation configuration. The operator can use this information, other operating conditions not explicitly considered in the optimization, and his/her own experience to select the most reliable TLS actions at each level of the tree for implementation. It is important to mention that the suggested switching actions will be a temporary solution to recover from the critical contingencies in a timely manner and the switched transmission lines would be returned back to service when the contingency is permanently mitigated.