Buy, Burn and Bash

The equipment is going to wear out anyway. This is the core concept driving ownership and operating decisions at any organization taking this approach; a company can do preventive and service maintenance on some parts of a transformer, breaker, or other electrical component, but there are other characteristics critical to its operation in which deterioration cannot be well controlled. Those failure modes will lead to its failure anyway, so why bother with servicing those other things?

For example, electrical bushings can be renewed and metal cases checked and repaired for rust, etc., but the electrical insulation in a transformer’s core will deteriorate with use and nothing short of completely rebuilding it will slow or stop that. In some types of breakers, contacts and mechanical parts can be renewed, but the tight, perfectly round and aligned metal castings supporting its trip mechanism will loosen, warp, and fatigue with every use; it is only a matter of time until it gives problems as a result.

Thus, this perspective holds, in both cases money spent on “proactive” replacement and renewal of components that can be repaired is largely wasted: modes of failure that cannot really be addressed will lead to the equipment’s eventual demise anyway. Better just to buy equipment that is reasonably durable, take measures to make certain nothing in its use contributes to really premature failure, and “burn it up” – use it until it fails – and then dispose of it and install another.

In some cases, this perspective is quite correct based on the facts, as will be discussed below, but in some cases a company takes this approach because it views the cost and effort required for proactive maintenance and care, and even routine service, as a distraction from its “core mission,” and/or because it has used this approach long enough that it knows no other.

The Standards-Driven Ownership Policy

Ultimately you can’t tell in advance exactly which units need service and which don’t. This is the key concept at the heart of a standards-based approach: a dozen medium voltage circuit breakers might look exactly the same from the outside; but one cannot tell until each unit is disassembled which will need service and which will not. Therefore, if owners expect dependable service they will have to service all twelve even if only four really are in need of that service.

This ownership approach manages operation, inspection, testing, service, maintenance, and refurbishment by sets of rules. Breakers of a certain type must be applied within a set of tightly prescribed operating guidelines and limits. Inspection will be done every XX years and those results followed up according to written rules carefully defining what is to be done or not done with regard to what is found. Every MM number of operations, or YY years, these breakers will be disassembled and serviced. After ZZ years, or when FF number of breakdowns or failures to operate have occurred, these tests will be done. If results are within a certain range, the unit will be sent for a factory rebuild. And so forth.

For example, based on the engineering and design studies that went into a high-voltage particular breaker, a manufacturer might recommend that the breaker contacts and trip mechanism be inspected and serviced after every six maximum fault-clearing operations. Over time, one particular owner might determine that for the set of two hundred such breakers that it owns, that service can be done after every ten cycles. Perhaps its system has lower fault duties and thus operational stress is lower, or perhaps the manufacturer’s recommendations were very conservative. Regardless, based on its experience, and its study and evaluation of past “as-found,” it has concluded that the longer inspection cycle will not cause problems, and over time that has proved to be the case. The rule becomes “inspection and fix as required after ten operations.” On the other hand, experience and “bad luck” may have led another company using this approach to determine that service after only five operations is required if it is to have the dependability of operation it expects from the set it owns.

Both companies manage maintenance of these breakers via a set of rules that specify when, where, and how that service will be done; the rules just differ slightly. Each company will have similar of standards for all other routine and periodic ownership, operating, and service decisions it must make with respect to breakers and allied equipment (relays, wiring conduits, etc.), each determined so that, if followed, the desired results, such as dependable operation, etc., are highly likely to occur.

Similarly, rules are established in a similar manner for the service of transformers, and for capacitor banks, and line regulators, and all other equipment. Electrical standards are established that set load and voltage limits on a particular type of transformer so that “premature” failures due to overheating and loss of life or overvoltage are extremely unlikely to occur. Voltage withstand levels for electrical cable will be set high enough that even as that insulation steadily degrades over time in service, it remains sufficient to avoid flashovers and failure for many years in service, and inspection of those cables will be specified to the extent determined to be necessary to keep failure rates below “unsatisfactory” levels. Intrusive inspections (disassembly, etc.) will be called on major equipment at frequent enough intervals to catch problems before experience has indicated they will get out of hand. Case and tank inspection and maintenance standards will call for inspection and re-painting on a certain schedule, all to avoid “premature” rust-through. And so forth.

These rules – standards -- are developed over time and may evolve as conditions change. When a technology or equipment type or situation is new, a combination of manufacturer’s recommendations, engineering assessment, and past experience thought relevant might be used to set rules. The standards might evolve rapidly as experience is gained and the organization “learns by doing.” A company may share and compare the standards it develops with other similar companies, participating in “standards groups” where its personnel meet with like personnel from other companies to discuss best standards with regard to use, service, replacement, etc. These groups might issue “industry standards” which reflect the consensus of all participants on what is recommended in general, but the company will modify them as it thinks fits its needs best.

Companies practicing this approach manage the results they get from their equipment by managing their standards. Generally, only when the ownership experience is not satisfactory will the standards be changed; if the equipment’s lifetime appears to be too short (e.g., MV cable is failing in large amounts after only 12-15 years in the ground), if breakdowns occur too often (breakers stick), if durability and dependability are inadequate (poles appear to be deteriorating too quickly), if safety becomes an issue (lines are failing on the ground more than expected) – the organization will study the causes and change the rules for either the specification of equipment, its utilization, or its maintenance, until it gets what it wants. Thus, over time, the organization’s standards evolve until the expected level of results – equipment lifetimes, failure rates, and operating costs, are satisfactory. Thereafter, as long as there are no unsatisfactory results, the standards do not change.

Condition-based standards-driven approach. Many modern practitioners of the standards-driven approach partition otherwise similar equipment into sub-categories based on diagnostics (e.g., DGA tests for transformers) or other critical factors (years in service) and vary the rules for operation and service in each sub-category (older units are not permitted to be as highly stressed, and/or are inspected more often).

This sub-categorization-by-condition-or-age approach is to a great extent driven by the aging infrastructures in the electric utility industry. As large numbers of equipment reach an age where the wear and tear of long service caused them to need more and different types of maintenance and refurbishment, the benefits of testing to determine condition are seen as part of the solution to a situations where results (failures, breakdowns) were getting out of hand: change the rules but only for a sub-set found to need the revised procedures.

Reliability-Centered Maintenance (RCM)

It is possible to identify what equipment needs service, and to focus attention on those units where service and repair will do the most good. This is the central tenet around which RCM is centered. “Reliability-Centered Maintenance” actually includes several slightly different maintenance prioritization methods that the authors have lumped together because they are based upon this central tenet. But all base their method on this fundamental assumption (usually true) that condition prioritization works. There are three variations that merely differ in how the organization executes the details of selecting the equipment for attention.

Condition-Triggered Maintenance

In this first approach of the three, maintenance is done on equipment based on measured equipment and performance related factors as compared to trigger points or thresholds. As examples, with respect to circuit breakers, this might be:

Number of operations: breakers are not maintained on a routine schedule every few years, but always after they have performed PPP fault clearing operations, whether that is only six months or three years.

Results of a timing test: the service and maintenance a breaker receives depends on the results of a timing test. If the unit fails badly it is rebuilt or full serviced. If it fails only slightly it is lubed, adjusted, etc., until it passes the test. Breakers that pass a timing test receive no service.

Similar rules would apply to other switches and mechanical equipment such as reclosers, load tap changers, capacitor switches, etc. For transformers, the rules might by based upon dissolved gas test results, or upon records of load (units loaded to emergency rating for more than two hours are inspected and serviced, etc.). For structures, there would be other rules, and so forth.

The types of data used to determine if maintenance is done, and the rules that trigger decisions to do or not do maintenance, vary a good deal due to the great variety of equipment types in use in any large electric system and the different companies making decisions as to how to implement RCM. Regardless, the common elements of condition-triggered maintenance are:

1) Performance or recent operating results, or test results for present performance or condition of the device are used to make the decision.

2) Comparison to a threshold or rule that determines if the unit will be serviced.

In the authors’ view this approach is not a full RCM or CCM approach – it straddles the boundary between standards-driven and the full RCM methods to be discussed below.

Standards triggered. This method depends on rules to determine maintenance. They are condition-based, but they are still rules, applied equivalently to all equipment of a type. Their contrast with other trigger methods will be highlighted later in this section when those methods are presented.

Condition-based. The method distinguishes equipment based on actual data related to each unit. In this sense it represents a tremendous change from the traditional standards-driven approach, in that it recognizes, acknowledges, and acts upon the fact that individual equipment of the same type (e.g, MD-80 breakers) and situation (old, heavily used) could nonetheless have very different characteristics and maintenance needs. Over time, this change alone can and will have a measurable effect on the “culture” of a maintenance department (See Section 15.3): this shift in group instinct and thinking is often as important as the actual improvements brought about by use of this method.

“Before” condition only. Decisions on service and maintenance are triggered only on the basis of condition prior to maintenance: “This breaker has operated six times so it will be serviced,” or “The transformer is gassing above the threshold rate so it will be torn down for inspection and rebuild,” etc.

Condition Models and Prioritization

By contrast, other categories of RCM include three aspects not a part of condition-triggered maintenance approaches. These are:

“Before and After.” Decisions on service and maintenance consider the impact maintenance will make on the condition or reliability by comparing status before to that expected after : “Based on its condition now the unit will contribute an estimated 129.8 customer minutes SAIDI next year, but if we do the service that drops to 417.6 minutes. The difference justifies the service cost,” or, “This life-extension service typically extends expected lifetime of cable in this situation from another 4.5 years to 13.0 years. The PW value and the lessened likelihood of an outage affecting SAIDI justify the cost.”

Model based. Condition-based and reliability-based maintenance methods typically use a model (although the organization might not call it that) of condition or reliability, to combine several measured data or tracking factors into a single-valued measure or classification of benefit or condition for use in determining service and maintenance decisions. Examples of this metric, used as the benefit in B/C computations, are:

Condition class. Dissolved gas test results, unit age, measured levels of audio and RF noise generation and the effects of possible service are combined into a measure of transformer condition: Near-new, Good, Fair, Poor.

Availability or unit reliability. Dissolved gas test results, unit age, measured levels of audio and RF noise generation, expected loading levels, and the effects of possible service are combined into a measure of expected availability or lack of it.

Service reliability or impact. Dissolved gas test results, unit age, measured levels of audio and RF noise generation, expected loading levels, number of customers served and factors describing contingency capability of the system in and around the unit, and the effects of the possible service are combined to estimate the customer service reliability for the next year.

Prioritization: Finally, true CCM and RCM methods prioritize potential or candidate projects and programs on the basis of benefit/cost, ranking them from most effective to least. Chapter 14 showed more fully how RCM methods work with detailed examples of both the models used to estimate before and after condition or reliability and the prioritization procedures used.

Once projects are ranked by benefit cost, the organization can approve and fund projects in order of effectiveness: if it has a very constrained maintenance budget this year, it picks only the “cherries” – those projects with very high benefit/cost. It thus gets as much “bang” for its limited bucks as it can afford. If it has a larger budget it picks projects farther down on the list – those that are best in the sense of what has not been selected so far – until the planned spending reaches the new, higher budget level. Prioritization does three critical things for the organization.

1) It maximizes the benefit it gets from the maintenance effort and spending it makes. Condition-trigger methods offer improvement over standards-driven results but do not maximize benefits.

2) It applies budget constraints in an orderly and effective manner so that the organization can spend what money it has most effectively.

3) It provides a means to evaluate the cost-effectiveness of the entire maintenance program and budget, which is why it is one part of the larger Asset Management approach, as will be discussed later.

Condition-Based Maintenance

Within this sub-category of RCM there are two further variations on the RCM theme, differing with respect to exactly how maintenance and service resources and monies are then allocated in one of two ways:

Maintain all units within a band of “good condition,” servicing them when they are estimated to be at the bottom of that range with a goal of bringing them up to the top of the range (but not perhaps to a level “as good as new”).

Maximize effectiveness of maintenance and service. Here, resources are added to keep units above some minimum threshold (“good enough”) but not necessarily to bring them all the way to some “top condition” level. Instead, the goal is to maximize the cost effectiveness of the entire program: if $1000 of service, re-painting, etc., will “buy” three years of further durability in operation for a particular breaker, at a cost of $333/year, and $2,000 would buy five years (a cost of $400/year), the organization will opt for the more frequent but lower cost program, etc.

In both cases, the organization uses a “model” to consolidate and combine multiple data on equipment condition or status into a classification, numerical condition measure, or other factor that is used to determine the merits of potential service and maintenance. For each potential project (i.e., do full service on breaker EA-36M) this is used to compute the “bang for the buck.”

Merit = (Condition measure before - Condition measure after)/Cost

This is computed for every candidate project being considered. The projects are then sorted highest to lowest. By picking the project at the top and then working down the list, approving the next, and the next, etc., until the permitted budget is exhausted, management can select the combination of projects that maximizes condition improvement. Constraints: The ability to handle constraints is important, such as “We only have crews to do 16 full breaker rebuilds this year,” or “If we do Project P, we must then also do Project Q.” “Spreadsheet” CBM ranking methods can handle a few constraints, but the best CBM methods use formal optimization algorithms to handle situations where the number of constraints outnumber the candidates involved.¹

Aging infrastructures are explicitly addressed with this method by making age one of the factors going into the condition model. Doing so effectively means having a statistical base of analysis to develop formulae for [age+other factors]Æ condition estimates. This will be discussed in more detail in Chapter 16’s examples. This approach is sometimes called reliability-centered maintenance: the condition measure (figure of merit in the equation above) being the expected availability (reliability) of the equipment during the next year or several years. However, the authors reserve that term for the method described next.

Reliability Centered Maintenance

This method is basically like the CBM method explained above, except that the “reliability” being optimized is reliability of service as seen by the utility customer or the industrial plant process. To do so, RCM uses additional data not used in the CBM process, information on equipment’s place in the system, the number of customers it serves and its criticality to their service, and the restoration time for a failure, etc. For example, suppose a utility can do maintenance on only one breaker but has two candidates for service. Both are in equivalent condition. Unit A serves 500 homes, is on a line that historically has had 3 faults every 5 years, and has an expected restoration time of 3 hours. Unit B serves 800 homes, on a line that is expected to have events 7 times in every five years, but where restoration time is only 90 minutes. A unit that has not been serviced has a 5% chance of misoperation, whereas a unit that has been serviced is evaluated as having only a 2% likelihood.

¹ One recent CBM project in which the authors provided assistance involved slightly less than 940 potential maintenance and service project, but over 2,700 constraints.

Reliability gain from servicing A:

500 homes x 3 events/5 years x 3.0 hours x (5%-2%) = 27.0 cust-hrs/year gained

Reliability gain from servicing B:

800 homes x 4 events/5 years x 1.5 hours x (5%-2%) = 28.8 cust-hrs/year gained

Unit B will get the service because it will yield a 7% greater impact in expected reliability improvement. All available resources for breaker maintenance and service will be allocated among breakers in a similar manner.

Furthermore, the organization will perform such “reliability gained” optimization to determine what it should service, and how comprehensively while comparing the cost-effectiveness of servicing breakers in this way to that of inspecting and pro-actively replacing cable, etc. Money and resources will be allocated between these two service areas, and to the inspection, testing, service and refurbishment of other equipment types, based on this type of analysis. Beyond that, the total budget for maintenance and service of all electrical equipment might be determined from this analysis, as the minimum needed to maintain reliability at a certain level. This type of optimization is often referred to as Pareto optimization and is discussed in much more detail elsewhere (see Chapters 5 and 12).

What CBM and RCM Do That Other Approaches Don’t

First, an organization following this approach believes that it is possible to track and estimate the condition of equipment, and that it can determine in advance how continued service and various types of service and maintenance will affect condition in meaningful ways; it has confidence in the analysis behind RCM.

RCM organizations do more inspection and more testing than those following buy-burn-bash or standards-driven paradigms, maintain equipment logs and records, and estimate and track condition. They also “optimize” allocation of resources (the aforementioned Pareto optimization).

Where CBM and RCM are Popular

Among utilities and organizations that want to maximize the reliability of limited budgets. An organization practicing Buy-Burn-Bash and having no issues with affordability of the approach will seldom even consider shifting to RCM: its value system and culture sees no value in optimizing something (service of equipment) that is of little value anyway.

Similarly, an organization satisfied with the results it is getting from the use of its standards-driven approach will see no any reason to adopt CBM/RCM either. It is getting the results it wants: nothing is broken so don’t fix anything. However, if challenged by a limited O&M budget, it may consider moving to RCM as a way of getting its traditional level of satisfactory customer (if a utility) or plant (if an industry) reliability at a lower overall cost. CBM/CM also appeals to companies that want predictable and equal budgets from year to year. Often, a budget driven by standards will vary from year to year as maintenance needs change (in addition to just natural variations, more severe weather one year might do more long-term damage and require more repair the following year, etc.). Regardless, a standards-driven company has no way from within the standards-driven paradigm to reduce spending if needed. With CBM/RCM, a company’s financial planners can reserve a certain amount of money for maintenance each year and CBM/RCM can spend that “well,” getting the most from it. Furthermore, it can estimate in advance what impact the budget reduction will have on reliability, telling decision makers the consequences of their action.

ASSET MANAGEMENT

Our electrical equipment exists to support our business in attaining its goals. A true asset management organization operates as if all of its assets, including electrical equipment, are business tools whose sole purpose is to further the business goals of the organization. Electrical equipment does not exist to provide power, reliability, and voltage within standards. Those products of equipment ownership, and their quality or lack of it, are important only to the extent they help the organization in obtaining its fundamental goals. Electrical equipment exists to make money, if the organization is profit-focused, promote growth, if the organization wants to grow in one manner or another, or to help maximize the number of orphans who can be housed, fed, and educated, if it is a charity focused on that goal.

How Does Asset Management Work?

Basically, the concept behind CBM/RCM – evaluate options on the basis of benefit per dollar and buy the most cost effective ones – is extended to other important desired results of the organization, and not applied just to reliability. The optimization is applied to all the many disparate goals of the organization – reliability, safety, durability, electric performance, work force morale, a desire to save not spend money, etc. – are somehow combined and coordinated in a context that makes it all work. That, in a nutshell, is asset management. The extension to other qualities beyond reliability uses basically the same approach as for CBM/RCM.

1. The organization must define a metric, or metrics, that measures the desired goal: SAIDI for reliability, maybe lost-time-injuries for safety, number of low voltage and sag complaints for electrical performance, etc.

2. It must be able to evaluate any and every project or program it could spend money on with regard to its impact on every one of these metrics, or have confidence that it has no impact on those metrics it cannot measure.

3. Pareto optimization – ranking the options by “dollars per unit of desired result” – and selection of the highest rated opportunities in each category can again be applied to identify which projects are best.

Only there are complications when more than one result (reliability in CBM/RCM) is optimized simultaneously. These complications can be mastered but they require considerable expansion of the analysis and optimization process, and a lot more work.

Multi-Attribute Optimization

“Multi-attribute” optimization approaches address the following situation.

- There are several attributes – each a “result” the organization wants as much of as it can get or “bad results” that it wants to minimize or control.

- Metrics – measures of the result sought – can be defined and applied for each attribute. They may be very different types of measurement for the various attributes: customer-minutes lost for reliability, lost-time injuries for reliability, voltage variation for power quality, Miller Alienation Workplace Satisfaction scores for employee morale, and so forth.

- Any project or program can be evaluated with regard to all metrics. Most projects will score zero in some categories but many will have some score in several: a project might impact reliability, have some influence of safety, and affect future O&M costs, etc.

² Attributes differ from criteria in that the organization wants as much of an attribute as it can get but only a certain value of a criterion. A criterion is a limit or target that must be met but for which little or no additional value is given if it is exceeded. Voltage must be above 110 volts to provide good service. Paint must be 7 mils thick to have enough adhesion and wear resistance. A result is a criterion if more than that desired limit is not needed. By contrast, more reliability is always a good thing, better safety always to be sought after. They are attributes.

- The “best” budget is sought: a combination of projects that obtains more results across this set of attributes, as the organization sees it, than any other.

This chapter is not intended to discuss how one solves this problem mathematically – single and multi-attribute optimization methodology is discussed in greater detail in Chapters 5 and 12 and given as examples in Chapter 16’s case study. What is important here is the basic ownership philosophies and understanding that lead an organization to want to apply that asset management approach, and how those approaches interact with and shape those philosophies.

So the important point here is that this problem can be solved, after a fashion. “After a fashion,” means this is done mainly by “structuring the problem” so that it is both mathematically tractable, and the answers obtained are practical (i.e., the answer is useful). And that is done via the steps discussed below. Not all are used in exactly the same way in all cases – there are slightly different ways to do asset management. However, in general, the following steps are taken anytime “Asset Management” is used by an organization:

1. Money is the “Attribute of Attributes.” The asset management approach is centered around minimizing cost or maximizing profit or optimizing some other monetary metric such as risk of failing to meet business goals. There is no doubt that this is done in many organizations because ultimately money is as or more important than any other result. But more fundamentally this is done because money – cost – is the one common thread relating all the candidate projects and programs that could be done. Some projects affect only safety, some just reliability, some reliability and electric performance and nothing else, some only interact with workforce morale and do nothing for reliability. But each and every one of them has a cost.

Strictly speaking, money is an attribute: the authors have never seen an organization or person that does not want more money if it can get it or more savings if that can be achieved. The reader confused about the fact that “single attribute” optimization such as CBM/RCM involves two attributes, reliability and money, needs to keep in mind that terminology is not always scrupulously correct and that money is always treated as distinctly different. The term “single attribute” optimization almost invariably refers to optimizing some desired attribute, be it reliability, or safety, or public image, against the “attribute of attributes,” money. Asset Management always includes money, too, and in some sense, it is also more important than any of the others, at least to the process and the decision making, within which it is the basis for how evaluations and decisions are made.

2. Include Both Capital and O&M Costs. Multi-attribute optimization can be applied to only O&M spending, or to only capital spending. However, in so doing it will tend to optimize that budget at the expense of the other. Multi-attribute optimization of a utility company’s O&M budget will maximize the effectiveness of O&M monies spent, but it might do so in a way that has negative consequences for capital. For example, an O&M spending focus on reliability and safety might neglect to spend on durability, so a utility gains good customer service and safety but spends little on paint and service to extend equipment lifetime: future capital costs for replacement are increased. The authors consider that true “asset management” involves the simultaneous optimization of both capital and O&M spending, considering the interactions and effects of one to the other.

3. Study the Trade-offs Involved. The organization wants as much of every attribute as it can get: zero-accident safety, world-class reliability, infrastructure durability better than a medieval cathedral’s, etc. But an asset management approach will quickly show it that it cannot have everything it wants, but that it usually can have anything it wants: there are trade-offs involved. One either spends a lot of money, or accepts less of one results in order to get more of another. As one optimizes capital spending in one way, it has consequences on O&M spending in another way. And so forth.

Asset Management is favored by executive management in many industries, including the power industry, largely because it can quantitatively identify trade-off options the company faces in this manner. True, it helps the organization “optimize” results – it leads to spending the final O&M budget that it decided upon as well as possible. But executive management is often more interested in understanding its choices than in honing its operating plan to the last decimal place. In fact, in the authors’ experience, organizations often do not chose the mathematically optimum plan: executives may think that the math does not quite address all the intangible aspects of their goals. But understanding the options the company has, and how tradeoffs affect performance and cost, capital, and O&M, one desired result versus another, is the key to the executives making their decisions well.

So, for the investor owned utility, what will it cost to improve SAIDI from 78 to 60 minutes if spending is optimized to do so at the least possible cost? If the company does not have the additional money to afford that, what reductions in durability, in electrical performance, in safety, etc., would it have to accept to free up enough money to fund that reliability gain? What if it is unwilling to reduce safety and wants to “take it from” the remaining areas? Can that be done? What will the resulting impacts look like? When will they set in? How long will they last?

Similarly, for the charity organization focused on orphan care in third world countries, will buying twenty high-quality 150kW field generators and maintaining them “by the book” enable it to care for more orphans than buying 25 lower cost units and skimping on maintenance, or not? For how long? With what risk of falling short of the expected goal?

This is the type of information executives in an asset management organizations want and use to make decisions with regard to their business, including their business’s involvement with electrical equipment, be that incidental or key to the company’s core mission.

4. Select a Plan that Maximizes Business Results. All of the above can be considered tactical evaluation and planning – analysis of how to maximize implementation and the options the company has in that implementation. The strategic part of asset management involves how the company decides what combination of trade-offs to accept, and why. There is seldom one clear best answer. A strategy that emphasizes durability (long-term lifetime) and “even business results” (steady year to year spending, low risk of failing to make profit targets) may emphasize different programs and spend money differently than one focused mainly on customer service reliability and system renewable, etc. All four aspects – durability, business results, reliability, and and system renewal – are important, and the organization can have more of one or two by trading away some of the others. It can have more of all four only by giving up more of the “attribute of attributes” – money.

Decisions at this level are always made by the people running the organization, and are basically strategy and policy and are made by the executive management of a company. Fundamentally, any organization, be it an electric utility or the charity, values electrical reliability, low O&M cost, safety, employee morale, and other reuslts because they contribute to its success. A particular plan is selected because those making the decisions – the company’s executive management – ultimately decide it has the greatest likelihood that it will result in the organization’s success, whatever that means to them. For an investor owned utility these might be operating profit, stock price, or executive bonus level. For a children’s charity this might be number of orphans cared for and public image leading to increased contributions next year.

Many companies, particularly large technically-inclined businesses, use the “output” of the electrical equipment asset management analysis as the input to a corporate model that those making policy and strategic decisions use to chart of the overall course of the organization. In many cases this “model” is a computerized numerical model that juggles investment, O&M costs and its emphasis, market share, pricing strategy, workforce size and skills, and investment in all these areas, against financial factors like cash flow, debt ratios, contingency needs, etc. In other cases the model is just the instinct and intuition of top executives, honed by years of experience and applied according to their judgment.

Either way, such models are used by the executives running organizations to evaluate their entire business, and answer essentially the same questions as asset management did for the electrical equipment: “What can I do and what can’t I do: what are the options and tradeoffs I have?” The information on the electric system and the asset management options that an organization has is only one set of inputs to its “corporate model” but often a very key one if electrical equipment is an important aspect of the organization’s function of needs.

Regardless, working from the information provided, the company executives pick a plan. That plan in turn rests on certain options from the electrical equipment asset management, which are now essentially “set in concrete” as part of the overall plan. For an electric utility, the selected plan might be something like “We will spend $511 million, for which we expect a SAIDI of 75 minutes, safety below .8 lost-time injuries per 100,000 hours labor, power quality leading to fewer than 25 complaints per 1,000,000 customers per year, a major equipment durability index of 47 years expected lifetime as we measure it, employee morale greater than 85 on the Miller Alienation workplace satisfaction scale. Details of how we get each of those and how they hang together were determined in the asset management analysis given to us. That plan and its recommendations are approved, and we expect the results promised to follow from that spending and effort.”

5. Set and Manage to Targets. The approved electric system asset management plan identified how much it would spend to be spent on each area such as equipment service, replacement, training, etc., and it identified each of the results that would be generated such as reliability, employee morale, etc.

This plan now permits the organization to pretty much de-couple the management of various aspects of its operations into single-attribute situations within each division or department. For example, suppose the asset management plan that has been approved called for $13 million for MV breaker maintenance. That budget can be assigned to Substation Maintenance for MV breaker maintenance, and within that, it has only to focus on doing projects that are most cost effective – using the money well. This is a bit of an oversimplification (breaker maintenance involves following safety practices in order to achieve the companies goals), but illustrates the point: the utility manages its overall combination of results it wants by studying the trade-offs and then assigning “vertical” responsibility among various groups in a way that makes each of their assignments pretty close to a single attribute management focus.

And so ultimately, the business sets its tactics based on the strategy it selects. But the process it took to select a strategy took into account an evaluation of the options it had with respect to tactics. An example case later in this chapter will discuss how this works and give diagrams of the process used at one asset management company.

Regardless, good executive management will define and select a strategy it considers is best for the organization. This will vary greatly from company to company: an investor owned business seeks to maximize profit - owning electric equipment doesn’t change that. A non-profit organization might define its goals very differently, as something along the lines of “We want to maximize the number of orphaned children we feed, keep healthy, and school.”

Typically, an organization’s strategy will not be to maximize that results it might achieve, but to minimize the risk of failing to meet targets for those results that it promises to meet. Executives running an investor owned utility might reject a plan that promises them a good chance of making 12% profit, in favor of a plan that gives them a near certainty of making 11.5%, etc. “Taking into account factors we cannot control like the weather and economy, we have set our profit goal at $290 million, a figure we are confident we will make. We have set reliability, employee morale, and other goals in a way that is most compatible with both achieving this goal and in being flexible to help us respond if conditions change, so we still attain that goal.”

Is there a “right way” to operate?

Perhaps the most fundamental contrast among these four operating approaches is with regard to whether each considers that there is a “right way” to practice equipment ownership and operation. As different as buy-burn-bash and the standards-driven approaches are, when implemented in an organization, either leads to a definite view of how equipment should be used and what service maintenance should be done: there is a right answer. For buy-burn-bash, this is “use it up and don’t take care of it.” The standards-driven approach holds that the right way to do things is “by the book” and views anything more as unnecessary and anything less as “lowering our standards.” Over time, organizations that practice one or the other of these two approaches develop a culture that has great confidence in the approach, that thinks and acts from this viewpoint instinctively, and that tends to regard deviations from “the right way” with cynicism, often as if any such proposal is near heresy.

By contrast, both CBM/RCM and the Asset Management approach never imply or lead to a conclusion that there is a necessarily absolutely correct way to do things in all situations, although they will identify a best way to do things in any one situation. Use of these approaches tends to put the fact that there are always compromises and tradeoffs into stark view, forcing an organization and its members to never forget that anything gained comes at the expense of something else given up. This is a fundamental difference that eventually leads to very different cultures inside companies practicing these different ownership approaches.

Buy-burn-bash and standards-driven organizations are in many cases strongly convinced they are not only “doing things right” but that there is no reason to consider improvement or change: “It isn’t broken, so don’t fix it.” There is often little internal willingness to review and refine, to improve, or to innovate, and these organizations often have a difficult time adjusting to external changes. Such organizations tend to operate very efficiently, and to make few mistakes with regard to their approach. But they can become complacent.

CBM/RCM and Asset Management organizations are confronted daily with the fact that there are different “ways to do things” and a range of tradeoffs, and that any decision has potential advantages and disadvantages versus the others.

Over time, these organizations develop a culture of constantly evaluating, even challenging the way they do things. Such organizations tend to be slightly unsatisfied with the results in the best of times. In the authors’ opinion, while the buy-burn-bash and standards-driven organizations might be much more confident and comfortable places to work, CBM/RCM and Asset Management based organizations can be much more effective, particularly in the face of change.

Missions and Cultures

An organization’s culture grows out of the shared experiences of its members, the type of work it does and how it does it, feedback from its successes and failures, and the leadership and influence of its management. It is fashioned by the routine work procedures and priorities established in the workplace, and the belief people and teams invariably develop that their work is important, that they are doing it the right way. Thus, cultures and basic work philosophies/approaches tend to support and strengthen each other in a type of feedback: given that it sees itself as successful, the culture will reinforce its own commitment to the use of the particular set of values, beliefs and methodology, and conversely, the use of that “successful” methodology will strengthen the culture’s belief in it and itself.

Consider for example two hypothetical organizations, both responsible for the maintenance and care of a T&D system and its equipment.

1) The first is an organization that has long used a pure standards-driven approach. Over time, given that the organization is successful (as it defines that) this led to a culture comfortable with, believing in, and adapted to the standards-driven approach: do things “by the book.” The culture will expect to have written rules and procedures to follow at every turn, and consider that adherence to standards and completion of work to prescribed requirements constitutes in some sense the end of its obligations.

This organization’s culture will de-value methods or approaches that analyze and tailor solutions to individual situations or equipment units: after many years with this approach the organization not only does not have the tools to take that approach, its members believe such an approach will not work technically or practically: that it is wrong. Instead, it firmly believes that performance issues, be they equipment operation or high failure rates or reliability issues, are solved looking at aggregate results with detail and then changing the rules or procedures to bring results back to an overall acceptable level. In short, it will believe that there is a right way to do things, period, and that that is prescribed by its standards.

When confronted with a sudden problem or unexpected surprise the organization’s instinctive reaction is to search within the organization’s standards-driven approach for a solution: what do the rules say? It will not think of or trust other approaches.

2) By contrast, suppose the second organization has for a long time applied a pure Reliability-Centered Maintenance approach to do what is essentially the same job as the first. Again, assuming it has been successful as it defines success, its members will grow experienced with, respect, and trust the CBM/RCM method. They will believe in methods that track inspection and test results on, that estimate condition of, and that prioritize work activities for individual equipment units or projects. Performance issues, be they equipment operation or high failure rates or reliability issues, are solved by gathering data on individual units and events, by looking at and classifying individual results, and then ranking or prioritizing or otherwise selecting items for action.

This culture “knows” that there is never one right way; that decisions on maintenance, and in a way everything and anything else, depend on background and other factors, that aggregate performance results can always be obtained by different sets of individual decisions and projects, and that all projects are in a type of competition with one another for priority. When confronted with a sudden problem or unexpected surprise the organization’s instinctive reaction is to first measure the problem, evaluate its implications (usually with a focus on reliability), and then structure a decision based on maximizing value.

These organizations might have the same purpose, but they are fundamentally very different otherwise. First, regardless of culture, they have very different capabilities at their disposal; neither group could tackle a challenge by applying the other group’s methodology, even if it wanted to – it does not have the tools, methods, or skills.

But beyond that, neither group would instinctively or even intellectually consider using the other’s approach. People in a standards-driven organization believe in methods that look, perhaps in great detail, at everything in aggregate and in the big picture. They seek to identify and apply broad rules and generalizations that will work in the management of groups of equipment: if 19 out of 93 MV breakers have suddenly experienced a certain problem, the solution is to find a procedure or solution that can be applied as a rule to all 93 – because the problem is probably just as likely to occur to the other 74.

People in the CBM/RCM organization believe in very nearly the exact opposite: look at and treat every unit of equipment as potentially different and apply resources in a focused, not generalized way. If 19 out of 93 MV breakers have had a problem, find out how to inspect and test MV breakers in order to identify this condition anomaly, then evaluate and prioritize its seriousness so that resources can be applied as needed to units that appear to have the problem, and not to the others.

Culture Can Change

Cultures will change and evolve over time. This can happen gradually, and often haphazardly, when something in or around them is changed: at many utilities a change in methodology from standards-driven to CBM/RCM is first resisted by the existing culture, then warily accepted, and finally becomes the core of a new culture. This can take some time, and workplace results and business performance are not always “pretty” during the transition.

Cultural change can also be driven and directed through effort and focus: often when a new CEO comes into an organization, she or he will start by saying that they want to “change the culture.” They mean that the present organization is not capable of reacting and performing as they require, etc. The authors have seen culture change effected at several companies with remarkable results as far as improvements in productivity, system reliability, etc. In the authors’ view, the fact that cultures can be changed leading to palpable changes in performance is the very best proof that they exist.

15.5.1 Big States Electric – Asset Management

Company Profile: Big States Electric is an investor owned electric utility, the result of a series of merger-acquisitions over the preceding fifteen years in which Big State Electric absorbed four neighboring utilities. All five companies are able to trace their company’s heritage back to the beginning of the 20th century. The company operates now as six “independent” operating companies, each within a single state, serving a total of 4.9 million connected meters with a coincident peak hourly demand of 24 GW.

Electrical Equipment and System: Each of the four companies it has recently absorbed, and Big State Electric itself, were the result of previous mergers, some going back into the 1920s. As a result the company is a conglomerate of different voltage levels, design types and historical operating methodologies and history. It owns 9,902 miles of overhead and underground transmission operating at 500, 345, 230, 161, 138, 115, 69, 46 and 34.5 kV and 57,800 miles of distribution line comprising 4,734 feeders operating at voltages of 34.5, 25, 23, 22, 19, 13.8, 117.5, 11, 8, and 4.2 kV. BSE owns 87 purely transmission voltage switching stations and 918 distribution substations with 2690 transformers with an aggregate nameplate rating of 58 GW. It owns 1,134,007 service transformers with a total installed capacity of 45GW.

Ownerships and Operating Approach: BSE uses an Asset Management approach for both capital and O&M budgeting and prioritization of spending, having made this transition after the most recent merger. All five companies that merged into BSE were standards-driven organizations, and initially there was talk of leaving each company free to continue to work from its own traditional standards even as they became part of BSE. The rationalization for this was the fact that since each now operated under the jurisdiction of a different state regulator there was no problem if they each had different practices and standards. A senior manager observed that while this undoubtedly true, the main reason was that no one wanted to take on a messy and potentially contentious job of integrating standards and standards-departments across all the companies.

But in order to gain approval for the latest merger-acquisition, BSE had to promise synergies and savings amounting to nearly $60 million annually ($1 per month per customer). The use of common planning, operating, and maintenance practices, and the integration of those departments into centrally administered groups, were among a long list of savings areas BSE gave stockholders and regulators who had to approve the merger-acquisition, in order to get approval.

The company waited for three years after the merger to begin this task. The resulting effort to integrate standards and guidelines coincided with the appointment of a new President of BSE’s Delivery Company, to whom the six operating company Presidents reported. She drove the effort toward Asset Management, insisting that the resulting merged organization and all six operating companies use a uniform, integrated asset management approach linked to BSE’s corporate model and strategy. This was an rather clever move in one way, senior managers observed, for it meant no one company had to accept any other company’s standards – they all had to make a transition to something else.

BSE’s asset management is administered by an Asset Management Coordination group reporting directly to the SVP of Delivery Operations. They creat and maintain the asset management tools and process and coordinate and support its use within the six operating companies. Figure 15.4 shows how the system is implemented. It is hierarchical, both top down and bottom up. Each operating company inputs all capital projects and its maintenance and service programs and projects into templates for the company’s PEPT (Performance and Planning Tool). These are analyzed in aggregate and Pareto and other analysis results output which basically given the needs and options each company has: “We can do SAIDI = 75 minutes and such and such performance in other areas for $XX, or SAIDI = 80 minutes and so and so performance in other areas for $YY, or,” etc.

Figure 15.4 Structure of BSE’s asset management implementation.

Figure 15.5 Age distribution of BSE’s 1,215 power transformers.

Figure 15.6: Top: Normalized PW annual cost for the next 29 years of three asset management scenarios for management of BSE’s aging transformer fleet compared to the predicted cost of “business as usual” under its former standards-driven approach (St-D). Bottom: SAIDI contribution of the four transformer management programs. See text for details.

Asset Management results and the options from the six companies are then analyzed as a whole, in aggregate providing the “options” that BSE executive management looks at to set its strategy and policy, and to assign T&D capital and operating budgets for each company. A good deal of the effort at this level deals with constraints created by six different operating jurisdictions: the company has different promised reliability targets and rates in each state, and requirements to be able to show it is within agreed upon regulatory limits in how it operates and uses ratepayer money with each. Yet it must operate as one company and achieve the overall synergies and savings it promised. Executives pick a plan that fits BSE overall needs and also fits within legal and policy decisions for each state operating company: segments of this plan are basically the “marching orders” for the six individual operating companies. Each company takes its part of the overall plan and is then free to operate as it sees fit, although it is expected to provide the performance (attributes and spending limit) promised in its submitted asset management information: its executives and management compensation is based on achieving that goal.

BSE upper managers and particularly the staff of the Asset Management Coordination Group are very frank about the fact that initially the asset management initiative did not go well. Inevitably there was resistance to the change within each operating company, including many people who felt the company was “lowering its standards.” Over time, these issues subsided and by the third year of operation under the new approach had mostly disappeared.

But BSE also discovered that while its asset management data, tools and process worked they were not dependable in one sense – the process is accurate and useful only when making incremental changes. BSE discovered what many asset management theoreticians, consultants, and internal coordination groups do not want to accept, that in practice the method is only accurate incrementally. For example, suppose it has been spending 15,800 hours labor annually on medium voltage breaker maintenance and is seeing a mis-operation rate of .71%. A detailed analysis that determines that if it increases hours by 5% to 16,600 annually, misoperations will be reduced to .67% is probably fairly accurate. Analysis trying to determine what would happen if hours were increased or decreased by 25% is very likely not to be dependable.

Aging T&D Equipment Infrastructure: Figure 15.5 shows the distribution of ages of power transformer MVA in the BSE system. The newest unit is less than a year in service, the average is 37 years in service, and the oldest unit is thought to have been in service for 72 years (records going back to its acquisition do not exist). BSE has investigated aging and deterioration and their expected effects and created three scenarios that it is using to manage decisions about service, refurbishment, and replacement. No one, including the Asset Management Coordination group that worked with consultants and vendors to produce the study, and upper management, is confident the analysis is completely accurate. However, the analysis is transparent, all the data and conclusions traceable, and the methodology makes sense and seems credible. Company executives have decided to use it as the basis for their decisions in this regard, and have received no pushback from regulators.

Figure 15.6 shows BSE’s analysis results for how it should best manage its aging power transformer fleet. The top diagram shows the annual cost of replacements, refurbishment, repair, inspections, and testing for four scenarios. The bottom number shows the project SAIFI contribution of power transformer failures and operating problems. “St-D” is business as usual under its traditional standards-driven approach. Atl1 is a plan optimized with a constraint that it would never spend more than the standards-driven approach in any year. It reduces PW of future expenses by 2.1% and the time-weighted value of SAIFI over the next 29 years by 3.2%. BSE’s staff did it to see if and how “optimization” could improve the situation even if constrained by budgets. Alt 2 is a plan unconstrained as to when money is spent but constrained to the same total PW cost as Alt 1 –a study done to determine if “early spending” on life extension, etc., actually pays off with regard to PW results. It works down SAIFI to 9.1% less than the St-D case, so the answer is “yes.” Alt 3 is an optimum plan as far as BSE’s Asset Management Coordination group can determine. It reduces PW cost over the period by 4.2% and SAIFI by 13% as compared to the standards-driven case. Again, the staff, BSE’s management, and the executive making the decision were aware that these results were only approximate, but the company went with transformer management alternative 3.

15.5.2 Mid State Electric Company: Reliability-Centered Maintenance

Company Profile: MIDSEC, as it refers to itself, is an electric cooperative serving a 4,800 square mile service territory that was once rural but is now the suburban periphery of a large and growing metropolitan area. It sells power to 198,000 connected meters with a peak hourly demand of 1,087 MW.

Electric Equipment and System: MIDSEC owns 391 miles of 230kV, 138kV, and 69 kV transmission lines, 131 power transformers, 197 high voltage circuit breakers, 312 medium voltage breakers and 75 MVAR of high voltage capacitors located at 48 substations, in addition to 4129 miles of 117.47 kV feeders, 175 line voltage regulators, 412 capacitor banks on feeders, and 31092 service transformers in service. The average age of its transformers is 22 years; over half of its customer base has been added in the last two decades due to the high growth of the metropolis nearby. An aged infrastructure is not a problem for MIDSEC. However it has a portion of equipment that is old including two substations and five power transformers dating to the late 1940s.

Ownership and Operating Approach: MIDSEC practices a combination of reliability-centered maintenance and “modern” standards-driven maintenance in which it sub-categorizes some classes of equipment into “bins” based on age and diagnostics, applying different rules to each bin. It applies RCM to vegetation management, to the inspection, treatment and replacement of wooden structures on transmission lines, to all its 131 power transformers and to 105 of its service transformers (all those over 1000 kVA). It has a Transformer Load Management program that addresses the other 3,987 service transformers it owns, which it considers to be an RC process.

All other equipment is operated, inspected, tested, and serviced “by the book,” according to a set of standards evolved from REA operating and maintenance documents but modified through many years of operational fine tuning to MIDSEC’s needs. The company “sub-categorizes” breakers by age and loading and number of customers into groups that receive more or less frequent and more or less comprehensive service. Similarly, equipment, structures, and conductor on lines in heavily forested areas with high customer densities are inspected on a rolling five-year cycle rather than ten-year cycles as in other areas.

MIDSEC’s move to RCM, as at many utilities taking that approach, was initiated from within the Engineering and Operating division and not by top executive and financial management as is the case with many Asset Management initiatives. MIDSEC is one of the few electric cooperatives to participate significantly in IEEE equipment standards groups as well as NRECA bodies. Its VP of Engineering learned of RCM while working in such meetings earlier in her/his career. A new VP of Operations brought on in 1998 had a background in the nuclear navy and so was familiar with RCM approaches. As they implemented it at MIDSEC, RCM was mostly a technical initiative aimed at improving reliability of service. Beginning in 2001, RCM at MIDSEC focused on reliability-as opposed to condition-based maintenance. In fact, RCM began with application to vegetation management.

When RCM was implemented, the company also revised loading standards. Normal and emergency loading limits for transformers and cable were increased after a very comprehensive “thermal” analysis of equipment operating stress and loss of life that showed there would be no impact on reliability. MIDSEC’s Engineering and Operations department personnel consider this re-rating part of the company’s RCM plan even though, strictly speaking, it did not address maintenance per se. The loading analysis categorized 15 older transformers in a sub-category of reduced allowed loading. The average unit in this “low-load” category had an expected 11.6 year remaining lifetime based on IEEE standard Arrhenius formula calculations of expected insulation degradation. The reduced loading limits are aimed at an average 20 year remaining lifetime for these units. Load reductions were arranged by re-switching the feeder system to shift loading to newer units in the same or nearby substations.

Figure 15.7 shows the results from MIDSEC’s RCM program. Eventually spending on equipment T&D was reduced by over 11%, while SAIDI has improved by nearly 10%: MIDSEC’s standards-driven methodology had been delivering good results, but seems to have been spending much more than necessary.

MIDSEC’s executives are pleased with the program but view their current

Figure 15.7 Annual T&D system service and maintenance costs normalized to system size and corrected for inflation, etc. (solid dots), and SAIDI (open dots). Anomaly in reliability and spending in 1989-1994 was caused by a design flaw in primary voltage lightning arresters that were replaced in ’94-95. Begun in 2001, RCM permitted MIDSEC to reduce spending on equipment maintenance by over 11%. Within four years SAIDI results were clearly dropping from an average of 88 minutes to about 80 minutes. Dotted lines show MIDSEC projections of future performance (see text for details).

Figure 15.8 Data for ten years of MV breaker maintenance at five electric cooperatives. Breakers were put into “bins” by years in service 0-5, 5-10, etc. Top: annual cost per unit in each bin was computed from maintenance records. Dots show the average for all 10 years of the survey for each age group, line shows the range of the data. MIDSEC was surprised that there appeared to be an increase in costs after only 25 years but not that cost began to escalate at 40+. Broad line shows the curve MIDSEC used to represent the cost-versus-age relationship in subsequent studies. This operating and maintenance approach is an interim step in a gradual evolution toward a full asset management method. As technology, time, and resources permit, they expect to add more sub-categories to those equipments they manage with their sub-categorizing standards approach, and eventually to extend RCM to a wider range of equipment types, and then gradually add more attributes to the optimization.

Aging Infrastructures: As mentioned earlier, MIDSEC’s growth over the past 15-20 years means that it has a relatively new system, at least when viewed as a whole. The average transformer MVA in its system is 21 years old. However, 15% of its system equipment is over 40 years of age. In an effort to analyze aging effects quantitatively, MIDSEC shared data and analysis effort for a statistical analysis of equipment maintenance costs, condition, and failures on its system with five other cooperatives in its region. One conclusion of this study was that breakdowns were more frequent and maintenance cost slightly more expensive as units got older (Figure 15.8).

Qualitatively the escalating cost with age was not entirely unexpected, although MIDSEC was surprised that costs began to rise at age 25. It had expected a later but then faster escalation as a function of age. Regardless, as its breaker fleet ages, MIDSEC expects maintenance needs and costs for it to rise. For planning purposes it uses the broad line in Figure 15.6 as a guideline: it can estimate costs for, say, five years ahead by simply assuming all its current breakers are five years older and applying the cost escalation given by the curve. Similarly, studies on other equipment types lead to a like analysis of future costs. In this way, MIDSEC projects its RCM-based spending will gradually rise over time along the dotting line projection shown in Figure 15.5.

An unexpected consequence of RCM that concerns MIDSEC management is that, as its system equipment ages, RCM will respond by spending less on service and maintenance in some categories. Basically, RCM seeks to spend money where it does the most good from the standpoint of improving reliability – where can I buy the most SAIDI improvement per dollar? As equipment, for example the breakers depicted in Figure 15.6 age, they become less reliable, or at least less reliable unless more is spent on them. “Shopping for reliability” by buying breaker maintenance becomes less attractive than, say, buying it through vegetation management or some other program. Perhaps today the RCM optimization balances breaker maintenance and tree trimming in a 50/50 allocation of funds. Ten years from now, that might be 55/45 in favor of the tree trimming.⁴

This non-intuitive response of RCM to equipment aging is the main reason MIDSEC wants to move to an Asset Management approach. Basically, “what is broken” with its current RCM approach is the system’s inability to manage anything more than short-term reliability improvement. It can’t manage durability (long-term equipment value and reliability), too, and it can’t relate all this to the business overall. For this reason, MIDSEC is committed to eventually moving to Asset Management. The relatively new condition of its current system equipment, and the slow progression of aging and its effects means there is no crisis and it can do so in an orderly manner.

⁴ The interactions between aging, reliability, and RCM-type methods are quite complicated and will be covered in much greater detail in Chapters 5 and 12. RCM responds well to equipment aging in a tactical sense, by shifting spending focus toward older equipment where reliability improvement is needed, but its overall strategic response is as discussed here – as equipment ages and costs more to maintain, it “loses interest” in doing so.. All single attribute optimizations have this flaw in some measure: if a systemic change in the cost of the attribute they are optimizing occurs, they may not be able to deal with it well. This is one reason multi-attribute approaches like Asset Management are favored for strategic business planning.

15.5.3 International Bulk Petroleum Fluids – Buy-Burn-Bash

Company Profile: IBPF is a long-established “vertical” oil company into exploration, drilling, recovery, refining, wholesale and retail sales of aviation fuels, gasoline, diesel, and fuel oil, greases and lubricants, and various plastic bulk stock under a variety of brands. It operates 91 major petroleum refineries and 297 distribution centers and sub-refineries around the globe. Refining capacity is a major bottleneck in its operation and it puts considerable effort and focus into operating all at 100% output, 24 hours a day, 365 days per year.

Electrical Equipment and System: IBPF owns plant power systems at all of its sites with peak hourly demands that vary from 12 MVA to 397 MVA. In most cases it owns the site substation and buys power at transmission voltage. IBPF obtains equipment for each of its sites based on the electrical standards and power industry practices in the part of the world where it is located: ANSI in North American, IEC in Europe, and JIN in Asia. It takes this approach both because it sources equipment locally and because personnel hired for the plants will be more familiar with what is typical in their continent.

Typical load factor for the larger IBPF refineries is over 95%. Most of its larger sites do a “plant shut down” or schedule shut downs of parallel portions of each plant for one or two weeks per year, so it can do equipment updating and maintenance of refinery equipment and systems. The rest of the time the plant is running at close to full capacity and loading is nearly constant. Downtime due to power outages can exceed $250,000 hour and even short outages can create situations that require a full day or more to return to full production. In locales where power supply is an issue its plants will have their own generators. Regardless, IBPF plant electrical systems are designed to be “bulletproof” – with redundant systems anywhere a failure would interfere with plant operation and with redundant equipment anywhere individual unit failures would shut down a system.

Ownership and Operating Approach. As regards the electric power systems throughout its plants, IBPF mostly follows a Buy-Burn-Bash approach, with only a few exceptions. Given the variety of sites it owns, equipment types it buys, and local laws and regulations it must obey, IBPF has to permit a certain amount of flexibility in electric operations at each plant. It has a standard framework for electric operations across all its companies and sites, a set of standards that pretty much mandates a buy-burn-bash approach to everything except protective equipment, which is maintained fastidiously and for which the company pays premium prices for advanced technology including protective monitoring and replaying systems directly tied into its plant control systems so production can adjust to power system issues and maintain maximum output. Transformers, capacitors, inductors, and similar equipment not connected with protection are all treated as “consumable equipment,” a term used by several IBPF plant managers, one of whom said, “They’re like flashlight batteries – buy them, use them up, and just pop a new one in when the old one is used up.”

Transformers at IBPF plants are sized very close to nameplate rating; typically a unit will have a continuous loading at or just under its nameplate rating. Under these conditions they last an average of 11.3 years at major IBPF plants (Figure 15.9), a figure roughly in keeping with expectations based on IEEE Standard insulation deterioration insulation deterioration rates.⁵ IBPF has done studies on buying larger size units and loading them to lower utilization rates (e.g., buy a 40 MVA instead of 32 MVA unit to serve a 32 MVA load). Its studies show this is not cost effective, partly because to get an assured longer lifetime for the larger units it would have to perform maintenance and service on the units.

Figure 15.9 Data on transformer failures versus time in service for three IBPF plants built in the late 1970s and early 1980s shows cycles with a periodicity of the average expected lifetime of the transformers in the type of service IFBP uses them to provide. Dots indicate actual data while the dotted line shows the statistical fitted trend over time. Most of these units are loaded to an average of 98% of nameplate for 50 weeks per year and have a calculated expected lifetime of 11.3 years. By year 25 many of the first series of replacements have also failed and been replaced. Over time the oscillations shown will gradually even out due to random variations in when individual units actually fail.

Breakers and their control systems at IBPF plants are its one exception to the company’s buy-burn-bash operating philosophy. They are serviced regularly and fastidiously and are replaced for one reason or another (often due to plant upgrades or reasons other than age) long before they are “old” – an average of every 22 years.

Within IBPF, MCCs (Motor Control Centers) and other local controls and operating systems, as well as the plant process control systems themselves, are the responsibility of and maintained by a separate group: Process Control and Operations which does not necessarily follow a buy-burn-bash approach with that equipment. Regardless, this equipment is usually updated on average every six to eight years as part of ongoing technology improvement of plant systems and thus never “age.”

Aging Infrastructure Issues. IBPF has no problems caused by aging or aged electrical infrastructure. Its plant managers and the executives in charge of its electric systems believe that it never will.⁶ The company recently decided not to pursue a new electric design-operation-management paradigm that it went to a great deal of effort to develop. Working with both electrical and plant design consultants and a software system integrator, it came up with an innovative new plant design that eliminated much of the redundant systems and equipment used in its previous plants, but assured equivalent production reliability by use of real-time incipient failure detection and “self-healing” control systems. This approach was extended completely through the electric system, resulting in a 61% reduction in its cost.⁷ Although the new plant design saved a good deal of money, the complexity and untested nature of the new system was ultimately too much risk for IBPF’s management.⁸

⁵ Fully loaded to its “nameplate” rating for all 8,760 hours per year, some power transformers deteriorate to 50% of insulation mechanical strength in less than 7.5 years.

⁶ The company does have issues with structural and chemical processing related infrastructures related to some of its older refineries.

⁷ Elimination of nearly one-half of the equipment reduced price to roughly half. But elimination of double buses meant tie breakers were eliminated, too. Similarly, elimination of one of each pair of redundant transformers reduced fault duties significantly, meaning lower capacity and lower cost breakers could be used, etc.

⁸ One or the authors helped with the risk assessment. Some IBPF plants pay for their capital cost in 17 months. While a savings of say, 10% is significant, it represents just 7 weeks’ production. A less than perfect “fail-safe” system could easily create many more weeks of lost production over the next 30 years the plant might be in operation.

15.5.4 Big State University System – Standards Based

Company Profile: Big State University System operates six campuses with a total of 396 buildings, an oceanographic research center on the coast, an inland agricultural research facility, and a full-service, 100 room hotel off campus for the School of Hotel Management. Total full- and part-time enrollment is 52,000. The school has 19,000 staff.

Electrical Equipment and System: Each of the BSU system sites is served by the utility in their area through utility-owned substations, with the university system taking possession of the power at primary voltage at the point of metering as it leaves the utility property. Campus power systems are 2kV, 4.2 kV, and 117.5kV systems, and mostly underground cable in steam tunnels and ducts and spot networks in basement vaults. Most buildings use three-phase 208 volt systems. The two research centers use some overhead 120/240 and 480 volt systems. The main campus chiller plant has two direct three-phase feeds at 117.5KV to power two 3,500 HP motors that run water chillers for the campus cooling system. Overall the BSU system owns 594 network or service transformers, 341 network protectors or equivalent, and over 40 miles of UG primary serving, 417 service points with peak demands from 20 kW to 4500 kW. Sum of non-coincident peaks at 417 delivery points is estimated at 156 MVA (some points are not metered).

The BSU system’s Electric Facilities Department is headquartered on its main campus in Capitol City. Its top management and engineering staff are located there, but the rest of its 171 personnel are distributed across all sites roughly in proportion to enrollment at each. The Director of Electric Facilities reports to the BSU Associate Vice President of Facilities and Services who reports to the BSU System Provost.

The Electric Facilities Department is responsible for specification, acquisition, installation, operation, and maintenance of all electric equipment at all eight BSU system sites, including building wiring and electrical equipment which requires slightly more than one-half the department’s budget and labor. In addition the Electric Facilities Department specifies and contracts for electric design and construction of facilities at new campuses and for new buildings on existing campuses.

Ownership and Operating Approach: BSU Electric Facilities Department is very nearly a pure standards-driven shop. The standards-driven approach has always worked well for it (and at many other universities), in part, the Director of Electric Facilities has observed, because it fits the mindset of the rest of the institution. Educational institutions are procedural and rule-based, using lesson plans, study guidebooks, periodic tests evaluated against written standards, etc., to specify and channel learning activities and structure their operations. Few types of organizations are more committed to doing things “by the book.” As a result, when the Electric Department manager meets with the Provost and other University system executives they understand, even feel comfortable with, the way the Electric Facilities Department operates and administers its responsibilities via a standards-based approach, and therefore have more confidence in and acceptance of its needs and recommendations.

The Electric Facilities Department’s Standards group maintains a set of guidelines for loading, for operational procedures like isolation of equipment for maintenance, switching and balancing load, and for maintenance of all equipment. These standards were based in part and follow almost completely those recommended by the American University Facilities Association committee on electrical standards, and are well proven as far as the Department is concerned. They have and will continue to slowly evolve over time. For example new OSHA Arc-Flash requirements implemented in late 2009 led to changes in BSU’s internal guidelines for inspection, testing, and maintenance of equipment.

BSU’s electrical facilities standards call for annual inspections of all equipment – a periodicity that many organizations, standards-driven or otherwise, would consider too frequent. In this case the motivation is safety – campuses are “schoolhouses” and a regard for the safety of the students and faculty is paramount. These inspections are primarily focused on safety but uncover a large “punch list” of correction items each year, providing all the routine examination of equipment BSU’s Electrical Facilities Department deems necessary.

One part of the BSU system is done with a completely different approach. In 1990-2000 the main campus was augmented by the addition of a 141-acre “Millennial Campus.” This provided expansion room for new engineering and chemistry graduate labs, a new physics department building and laboratories, but it was mainly a co-operative industry-academia site, where companies would build new R&D facilities to share resources and work. To attract high-tech companies to its shared resources Millennial Park, it is billed both as a next-generation IT environment wired with fiber and high tech systems throughout, and as a “power quality park” with near-perfect reliability and voltage stability of power delivery, and near-zero harmonics and voltage transients. The system is a modified distributed network using European ring-main buses and power electronics and storage throughout, along with a good deal of automation, all to maintain continuous power flow, stable voltage, unity power factor, and harmonic content near zero. This system is operated in a purely RCM manner, something that the director does not believe is necessary from an operating standpoint, but views as a marketing strategy to give further distinction to the area’s advantages to power quality conscious companies.

Aging Infrastructures While some of the BSU campuses are relatively new, having been built or greatly expanded in the last two decades, three date back to the early 20th century. BSU has a good deal of old electrical equipment of all types on these campuses. As discussed above, all equipment is inspected frequently. On the older campus underground cable runs through steam tunnels between buildings and the central plant and is therefore relatively easy to inspect along with equipment in vaults and building basements. BSU technicians visually scan all cable, splices, and junctions (three-wire splices) and scan with an infrared detector (inspections are typically done in summer when “steam” pipes are carrying chilled water and the tunnel are quite cool). This work catches a good deal of deterioration-related problems before they cause failures or cable vault fires, etc. BSU makes what it calls pro-active replacements before expected failure occurs, but still has considerable numbers of failures – electrical service is seldom interrupted because most of the systems are spot- and distributed-network secondary design.

Figure 15.10 shows failure-replacement rates for cable and splices and junctions (three way splices) for primary and secondary cable on BSU’s three oldest campuses over a thirty year period. There is no dependable data on equipment ages, etc. The data show failure-replacement rates have escalated rather linearly and by a total of 50% over three decades. BSU’s Director of Electric Facilities considers the cable failure escalation to be age related, but attributes part of the escalation in splices and junction failures to the poorer quality of materials and labor of the last twenty years. The data do show a higher rate of increase in failures since 1990, and a number of splices have been repaired several times, but there is no conclusive evidence proving this.

BSU’s Electric Facilities Department is most concerned about the age-related deterioration of the electric system civil facilities on these three older campuses. Some concrete and brick vaults, manholes, and tunnels date from just after WWI and while in good condition for their age, are deteriorated to the point they are a concern. A number of these rated worst condition by visual inspect were “cored” or otherwise sampled and will be evaluated by the Civil Engineering Department’s materials labs.

Figure 15.10 BSU’s 30-year record of underground electrical cable section failures and proactive replacements (solid dots) and splice and junction failures and replacements (unfilled dots) at its three oldest campuses.

Overall, BSUS’s Director of Electric Facilities and his staff acknowledge that at least some of their campus systems have aged infrastructures and aging-related failure and maintenance cost escalation problems. Overall maintenance costs have increased slightly faster than inflation and campus growth can explain for the past fifteen years, but budgets addressing this need have been approved every year, largely, the Director believes, because his department’s standards-driven approach is credible and respected, the Provost and board are familiar with the aging infrastructure problem, and his department has well-documented data on failures and maintenance costs to back up his budget requests. For this reason, BSU’s Electric Facilities Department plans no major changes in how it operates the system it owns.

15.5.5 Third-World Orphan Rescue Foundation: Uniquely Standards Based

Company Profile: The Third-World Orphan Rescue Foundation (TWORF) is a secular charity with a publicly stated goal to house, feed, and educate as many orphaned children in undeveloped countries as possible. The organization is headquartered in New York City and funded purely by support of philanthropic foundations and by individual donations. TWORF believes that by educating orphans it is not only improving their lives but indirectly the lives of others in those nations, by providing leaders and infrastructure development experts to help those countries develop. Its most recent annual report said it was caring for just over 19,000 orphans.

TWORF operates 43 “schools” – orphanages – in 21 countries. Almost all are located in rural areas and are basically small self-supporting villages, each with a population including orphans, TWORF personnel and indigenous staff, of about 600. Most orphanages have a “production” side; many farm the surrounding land for both their own food and to sell in local markets, others have a small factory or facility, such as manufacturing natural-fiber rope and netting, or making furniture – something that both earns money for the orphanage and teaches a trade to the older children.

Electric power is more central to TWORF’s success, and power systems equipment is more of a concern to the charity’s board and executive management, than an outsider might expect. Electricity provides many functions that TWORF considers essential to orphanage success: it powers water well pumps and conditioning equipment, provides refrigeration for food in its kitchens and medicines in its infirmaries, powers communications equipment including navigation beacons at landing strips in rural sites, runs televisions and computers for both education and recreation, powers equipment in its machine shops and the manufacturing equipment if any, at each site, and, of course, provides lighting and other conveniences. In addition, the teaching of electrical equipment service and repair as a vocational subject is a focal area for TWORF: electric systems are crucial to economic development and by training future electric system technicians it is enabling the countries where its orphanages are located to grow.

All forty-three TWORF orphanages have facilities to produce their own electric power and run what is essentially a private power system inside the orphanage grounds. Seven orphanages do have a connection to the local power grid, but 37 have no source of power but their own. As one TWORF executive put it, “Pretty much by definition, if local power supply was available and dependable these would not be third-world areas.”

Electric Equipment and Systems: TWORF owns a wide variety of electrical equipment types and systems with little uniformity in anything among its 43 orphanages. Peak power needs at its sites vary from just over 80 kW to 300 kW, with an average of about 210W per resident. It has both “European” (50-Hv 250 volt three-phase) and “American” (60-Hz, 120-1Ph/208 3-Phase) systems, and just about all types of equipment possible. A majority of generation is medium speed diesel but there is no uniformity in the types or equipment used, etc. Four sites use small steam generation fired by local fuels, including “agricultural refuse” at one. Some orphanages use “renewable energy” such as windmills for water pumping and passive solar heating for water heating, etc., where possible, but there has been little renewable energy until recently. This is changing as the organization has begun soliciting contributions to help it obtain wind and solar power. TWORF’s records indicate that it owns 228 diesel or fuel-oil powered combustion (piston) power generators of from 15 to 125 kW output, four steam-powered generators (two piston powered, one turbine, one sterling cycle), four small hydro plants powered by water wheels or “run of pipe” water turbines totaling just under 400 kW output, 87 wind turbines varying in size for 2 to 50 kW, and 23 PV “generator sets” averaging about 8 kW each. It currently has no micro or small gas turbines as the organization does not believe it has the skills to service them.

TWORF has three unbendable rules about power systems and equipment at its sites. First, the “power system” at each site and all electrical equipment operates only at utilization voltages below 260 V: no primary is allowed on site except in a fenced substation owned by the local utility, if there is one. Similarly by policy it does not deal with 440, 480, or 600 volt circuits and equipment either. Second, all circuits are overhead, and uniformly made with covered conductor insulated to 600 V. Third, for reasons of both safety and simplicity, there are no circuit breakers, only fuses.⁹ A fourth guideline calls for all circuits to be built and operated as three-phase open loops with isolation switching at each feed point, as is the fairly standard European LV design practice, but this is not mandatory, merely recommended.

Unique Ownership and Operating Approach: Management of electric systems at TWORF’s 43 orphanages is a “flexible and pragmatic” standards-based approach that unique as far as the authors are aware in one way: at every one of TWORF’s 43 orphanages, the power system is a vocational learning tool and the “Manager of Power Supply” is an instructor whose job description is to both run and maintain the site’s power equipment and train orphans over 15 who show mechanical aptitude to maintain and service electrical equipment. The two functions of system operation and instruction are heavily intertwined, not just in the manager’s assigned duties, but in TWORF’s design and administration of its electrical facilities and standards and in its maintenance and documentation processes organization-wide. The organization considers that “electric skills instruction” is one key area for educational development that particularly helps boost local economic growth and prosperity and thus focuses on driving this.

⁹ Building wiring systems might have breakers in their service panel boxes, etc., but incoming utility service, and all lines and equipment used to distribute power through each orphanage campus, are fused.

Ownership: TWORF’s New York based headquarters staff focuses on obtaining electrical equipment through donations or by purchase of used (often very well used) equipment from supporting foundations. It maintains records on what type of system (European or American) is at each site and lists of the equipment and requested needs from each. They try to address requests by school directors (“We need another 50 kW generator”) but take what they can get by donation and ship it to where it seems it will be most useful. TWORF headquarters buys equipment only as a last resort, preferring to let local School Directors, or regional Superintendents, buy equipment locally. Each School Director is pretty free to make any decisions on administration of the equipment, even to the point of selling or trading equipment that is obtained by headquarters and shipped to the orphanage, as he or she sees fit. Headquarters maintains an extensive library of service manuals and guidebooks on all types of electrical equipment, which it provides as needed to each site, trying to support what its management wants to do with respect to ownership.

Operation and maintenance, on the other hand, is tightly prescribed, with the three inviolate rules given earlier: power systems are exclusively utilization voltage with no primary permitted (except for incoming feed from the utility), all circuits are uniformly composed of covered conductor, and only fuses – no breakers – are permitted within the power system, although some individual building wiring systems have breakers in their service panels.

Beyond those rules, TWORF has very tightly documented standards for electric operation, service methods and procedures, and particularly for safety and work practices. It has a mandatory training program for power equipment operation and maintenance. This program and documentation is for all its maintenance and operations personnel, but is also used for instruction in its vocational programs in electric equipment service and repair. Many of its operations and maintenance personnel are the students themselves, who do such work as part of their training.

Annual inspections are made at each site by TWORF’s Superintendent of Water and Power or one of three assistants. Standards and inspections focus mostly on safety, completeness of required documentation, and the quality of curriculum, instruction and results in the vocational areas of water and power equipment and systems. Reliability of power supply, equipment condition, and even operating costs are factors not addressed at all, unless the School Director has issues with them.

With respect to safety, “Rule #1” is that all maintenance, repair, and service is done with the equipment shut down, de-energized, and grounded. There are no exceptions. Safety, including grounding and work practices, is a focus of much of the electric instruction program in the first year. Maintenance and service of equipment itself is incredibly resourceful: some sites make replacement parts for generators and re-wind transformers. Several perform such services for surrounding businesses and institutions. One site is in the process of “spinning off” an electric service company staffed completely by orphanage school graduates.

Aging Infrastructures: TWORF is unique among those the authors have worked with in not regarding aging power equipment as any kind of problem. The organization’s culture seems to value old but serviceable equipment as both a symbol that it does not waste resources and an advantage from the standpoint of on-the-job training: TWORF has a lot of student labor available in its vocational classes and so can handle a good volume of repair and service. Its records indicate that the average generator it owns was 20 years old when donated or obtained and is now 32 years. As mentioned earlier, its Water and Power Managers and their students are quite resourceful and fewer than 4% of its generators have been retired as unserviceable – once obtained they are kept in service “perpetually.”

Figure 15.11 shows analysis results for data from thirteen years of TWORF operating records covering a total of 318 motor generator sets in the 25 – 125 kW range. The data shown are used by TWORF’s central headquarters staff and local school administrators to estimate electric system budgets, etc., but the organization’s primary purpose in gathering it is to plan backwards from the hours of labor needed for this and similar electric-system related activities to determine the number of vocational students that can be given useful work and on-the-job training in this area of skill.

Only data on units owned and operated by the Foundation for two years or more havebeen used (many units donated to the school need major refurbishment and give problems in their first two years). The dashed line is the authors’ best third order statistical fit to data on the number of units found to be un-repairable as a function of age.¹⁰ The majority of failures that result in TWORF scrapping a unit for spare parts are mostly due to failure of major castings such as engine blocks, generator casings, crankshafts, etc. Non-availability of parts is not an issue: except for major castings the schools can make, or scrounge, most any part needed.

¹⁰ For both values plotted, generators were grouped into “bins” of 5 years of age: from 0-5 years, from 5-10 years, and so forth. Results were only significant for the range of ages from 15 – 65 years. Trends shown for newer and older generators were estimated by the authors, since TWORF has few generators less than 15 or older than 65 years.

The shaded area shows expected annual labor requirements for service per motor-generator and includes all systems in the unit, mechanical, electrical, and electronic or electromechanical controls. The plotted results are shaded to indicate something of the statistical range of variation in service requirements as units age. Hours of service required are measured in “apprentice” hours.^11,12 The most interesting point from the author’s perspective is the variation in hours/year of service required as units get older. Intuitively most managers and engineers associated with electric equipment will expect that as units get older they will vary more in maintenance and refurbishment needs, but this actual data that shows the very wide spread of “unpredictability” in service requirements that develops as units age, in addition to a steady increase in the average need. The standard deviation of service needs by age 50 is over 50% of the mean.

As a general rule, conductor, cable, switches and fused cutouts are bought new. Poles and structures are locally manufactured, usually by the school’s carpentry shops. The average age of this equipment is estimated at 16 years. Failures are not a reported problem at any of its sites. Documentation is quite good so it was possible to determine that over the whole organization, failure rate for circuits due to all causes other than external damage (truck backing into a pole, etc.) is .2 failures or repairs per year per mile. Similarly records indicate SAIDI is equivalent to 98%, with the majority of that being the occasional late fuel delivery causing cutbacks in power generation.

TWORF does think it has a “new technology infrastructure” problem, with power renewable generation it has tried to add in the last decade. Much of this is a result of unfamiliarity with the specification and purchase of equipment so that much of what was obtained is incompatible with the skills and service capabilities of its schools. All but 8 of its 87 wind turbines were new when installed and the average age of units now in place is only six years. All ran well for two to three years, but a majority now give problems within five years of installation: blade angle mechanicals and blade-to-generator gearing breakdown, and DC-AC power electronics often need service; both require tools and skills the schools do not have.

¹¹ Apprentices are students within one year of graduation from the vocational school and the students leading service and maintenance work under the tutelage of an instructor.

¹² Data used in this analysis did not take into account the size of units –a 125 kW unit typically takes somewhat longer to service than a 25 kW unit. However, the authors do not consider that to be an issue in these data plots’ validity. Data were checked and there is no strong correlation between age and size in the data, and the point of this Figure is only to demonstrate the effects of aging from actual data.

Figure 15.11 Data from TWORF records on motor-generators used at its orphanages around the world show the effects of age on annual labor required and % deemed unusable each year. See text for details.

TWORF has recently decided to use only singly-fed induction generators on its wind turbines to avoid these problems. Similarly, 30% of PV panels and inverters fail within three years. Diagnosis leads to no identified causes or systemic issues, but local shops cannot service the electrics or repair panels when PV cell laminates split, etc.

Overall, TWORF is committed to its present system and plans no changes other than to retire all its PV generation and all wind turbines that do not use induction generators, and shif to more use of sustainable energy in non-electric form.

15.5.6 Huey Longwaites International Airport – Buy, Burn, Bash

Company Profile: Huey Longwaites International Airport (LIA) is one of the nation’s newest airports, with a runway and ramp layout designed to minimize taxi time to and from terminals and promote fluid air traffic in and out of the airport at peak periods. Opened in October of 1994, the airport features “Smart” everything including automated people movers, baggage systems, and traffic light and control systems into and out of the airport. LIA is a 24 hour a day field with arrivals and departures internationally every hour of every day of the year.

Electric Equipment and System: The airport has a peak annual hourly demand of 72 MVA, with an 84% annual load factor, an indication of just how 24/7 its operations truly are. It owns the power distribution system within the airport territory boundaries, which is served by two “split ownership” substations, located at opposite ends of the airfield. Both are served by 230 kV transmission looped through the substation. Big State Electric owns the high-side equipment, including three 230/25 kV transformers at each site (two operating and one on hot standby), and the 25 kV buses and supporting equipment. LIA owns everything from the 25 kV feeder breakers on out to its system. In the first 15 years of operation, there was power outage to the airport two times for a total of five hours, both incidents caused by failure of the utility feed due to a systemic problem traced to protection coordination issues that occurred under unusual operation conditions of a nearby independent power provider.

LIA’s internal airport delivery system consists of 39 25kV/600-480-208V “substations” each a spot network, fed from 42 miles of mostly UG 25 kV primary feeder operated as closed loops with automated switching. It owns and operates a total of 138 25kV/utilization voltage transformers, with 131 in service and 7 kept in storage as spares. Within the airport system service reliability has been perfect at the terminal level: not one of the five airport terminal buildings or the control building have lost all power since the airport opened. Transformer annual peak demands are on average 96% of nameplate. Average annual load factor is 93%. Expected time to insulation half-strength calculated based on ambient loading and the Arrhenius formula is 16.4 years. However, this figure is considered optimistic by LIA’s electric department engineers, because it does not consider the rather high levels of harmonics in the system due to the heavy use of digital equipment and fluorescent lighting throughout the airport.

Ownership and Operating Approach: LIA practices something very close to a pure Buy-Burn-Bash approach to operation and maintenance, and its service staff and management firmly believe that all electrical equipment is expendable and that maintenance practices ultimately will make little difference in service lifetimes. By intention no inspections or maintenance other than routine annual visual checks for obvious leaks and safety hazards are done to any of the electrical equipment other than the breakers and protection systems. These are maintained under contract to an outside firm and are inspected annually and tested and serviced every three years.

Aging Infrastructure. Other that two early failures attributed to flaws in equipment as delivered, during the first 10 years of operation there was only one transformer failure and two 25 kV cable section failures. In early 2002 (8 years after opening) a full “security” inspection was carried out due to terrorism concerns after 9/11, at which point LIA installed equipment to monitor access to all facilities and track temperatures and conditions in each vault and duct, and “hardened” all data-com and control systems, etc. Since the additional cost was minor, visual inspection of all equipment and civil facilities for condition was added to this security inspection. This produced a set of over 600 items that could use attention. Electric department engineers “prioritized” this to a very short list that addressed several bad cases of case rust (including fixing the reason water was on the floor of two vaults), and replaced two sections of cable in an under-runway tunnel that had what looked like damaged sheathes. They also re-did several splices that infrared scanning indicated were hot. LIA records indicate that in the eight years since, the other equipment on this list has had failure rates no higher than other equipment at the airport.

Figure 15.12 Failures of transformers (top) and primary voltage cable sections (bottom) during the first 15 years of operation of the LIA airport electrical system.

In the seven years since this set of inspections, transformer, breaker, and cable failures have occurred more frequently and indicate expected average lifetimes of 14.8 years for transformers, 47 years for breakers, and 19.3 years for cable. Figure 15.12 shows the failure history and projected trend: these trends clearly indicate uneven and increasing expenses in some years but LIA management is not concerned. That has been anticipated and planned. LIA management and electric department personnel are satisfied with their electric system and its operation and maintenance practices and plan no changes.

Asset Management, if Done Well, Always Gets the Right Answer

In closing, the authors wish to note that this chapter, although focused on non-technical issues, revealed many of the limitations and issues that have led them to heavily favor the Asset Management approach. From an analytical and technical perspective, Asset Management, assuming it is well executed, will always determine what ownership and operating policies are best for an organization and particular situation. If buy-burn-bash is best, it will determine that, as well as any slight variations or variations or mixtures of that with other methods that will work best. If a standards-based approach works best, Asset Management will conclude and recommend that, if set up to assess that question. It will “dispatch” budgets in CBM or RCM approaches optimally. And of course, it will fulfill its role if executed as a full Asset Management program of gaining the best results possible with any budget, showing management its options, and computing the overall cost-effectiveness of the company’s spending plan.

That said, one must keep in mind that the data collection, clean up, filtering and aggregation needed, the condition tracking and analysis required, and the analysis and process management involved in Asset Management can be a considerable burden to any company owning significant amounts of electrical equipment. If the optimal answer is “buy-burn-bash” then that effort is largely wasted – the overriding reason that organizations using that or less-expensive equipment management approaches do no such analysis.

Technically, that argument is unassailable, except for this: the culture that Asset Management develops and creates and supports around it, of knowing there is no “right answer,” of seeing all decisions as compromises that gain and give at the same time, of using a common business basis to drive all evaluation and decisions, is far and away the best culture for any electric system owner. Organizations that “think” like this make fewer mistakes, come closer to optimizing results even when viewed in hindsight, and react far better and quicker to unexpected surprises and change.