Aging Distribution Infrastructure at Pacific Gas & Electric
Submitted to: Pacific Gas & Electric Company 123 Mission Street San Francisco, CA 94177-0001
Submitted by: Dr. Richard E. Brown KEMA Inc. 3801 Lake Boone Trail, Suite 200 Raleigh, NC 27607 919.593.2860
July 18, 2005
18-2
(PG&E-4)
EXECUTIVE SUMMARY KEMA has been retained by Pacific Gas and Electric Company (PG&E) to review its distribution aging infrastructure situation as it relates to the 2007 general rate case filing. Major findings of this review include: General. Aging infrastructure is becoming an increasing concern for the PG&E distribution system. As such, PG&E should not delay implementation of its proposed practices in the areas of substation equipment, underground cable, and wood poles. Proposed spending in these areas are conservative and justified. PG&E will increasingly have to spend money in these areas as the aging process continues to worsen the condition of equipment. Asset Management Processes. Asset management processes at PG&E are equal-to or better-than typical large investor-owned utilities in the U.S. This includes organization, project ranking, budgeting, and cross-program optimization. These processes are sufficient to make good decisions with regards to present aging infrastructure issues. Substation Equipment. PG&E is effective at identifying major pieces of substation equipment that are likely to fail. This is based on a mature Substation Asset Management program that has been in place since 1993. However, this program has not been able to replace all of the equipment that it recommends for replacement due to budget limitations, and increased spending is warranted. Underground Cables. PG&E can expect to see a dramatic increase in underground cable failures in the future. To prevent customer reliability from being significantly reduced, PG&E must eventually become much more aggressive in proactive cable replacement. In the short term, PG&E proposes to address specific problems, and to build a comprehensive database so that decisions related to a broader range of cable issues can be made in the future. This approach is conservative and justified. Wood Poles. PG&E has a substantial number of very old wood poles; to date, no negative trends related to aging have been identified. However, the number of pole replacements in recent years is unsustainable, even by conservative estimates. The PG&E proposed increase in pole replacements is conservative and justified. Back Tie Capacity. Older parts of the PG&E distribution system are heavily loaded, resulting in a limited ability to reconfigure the system and restore customers after a fault occurs (referred to as back tie capacity). PG&E has processes in place to justify reliability improvement spending based on cost-to-benefit analysis. PG&E needs to increase its back tie capacity on many feeders, and back-tie-capacity projects selected through these processes are justified.
18-3
(PG&E-4)
Distribution equipment at PG&E is getting older both in terms of average age and the percentage of very old equipment. The reasons for these trends are the same as for the industry as a whole: extensive construction activity in the 1960s and early 1970s, lower load growth since the mid1970s, and increased equipment loading since the mid-1980s. Although the situation at PG&E is worse than average due to the particular nature of the PG&E service territory (e.g., old development in San Francisco, mild climate), the effects of aging infrastructure have not had a large impact on customer reliability up to this point. However, there are recent signs of equipment deterioration, which means that PG&E increasingly will have to become more proactive in addressing issues related to aging equipment. Many other utilities are recognizing that aging distribution infrastructure will require increased levels of investment. This has resulted in a large increase in the number of recent U.S. rate case filings, and in the aggregate amounts requested in these filings. Often these increases are contested by ratepayer advocacy groups, which strive for adequate utility performance for the lowest possible rates. KEMA understands the need for ratepayer advocacy, but has the opinion that increases in distribution aging infrastructure spending at PG&E are required to both ensure adequate service and to minimize the life cycle cost of equipment ownership. With regards to aging infrastructure, PG&E is not perfect. The substation asset management program is best-in-class, but recommendations often are not fully executed. Cable replacement in recent years has been too low. Feeders are very heavily loaded, which will result in accelerated equipment deterioration and a limited ability to restore customers after an outage. These observations show that PG&E has historically made decisions tending towards lower spending, which is not necessarily bad since there must always be a balance between system performance and spending. However, PG&E’s historical approach is not sustainable since it will result in inadequate system performance and higher total cost of equipment ownership. The PG&E planning and budgeting processes are able to identify good spending decisions as they relate to aging distribution equipment. These processes have successfully identified areas requiring increased levels of spending, have successfully identified specific projects that should be implemented in the short term, and have successfully identified data that must be collected so that efficient decisions can continue to be made. This said, aging infrastructure issues at PG&E will get worse before they get better, and will require an increase in spending in the short, medium, and long term. This spending is required to reduce the life cycle cost of equipment ownership, to ensure safety, and to continue providing adequate levels of service to customers.
18-4
(PG&E-4)
TABLE OF CONTENTS EXECUTIVE SUMMARY ............................................................................................................ 3 1
INTRODUCTION .................................................................................................................. 6
2
OVERVIEW OF AGING INFRASTRUCTURE................................................................... 7 2.1 Equipment Aging ............................................................................................................ 7 2.2 Equipment Age Profiles.................................................................................................. 8 2.3 Population Aging Behavior............................................................................................. 9 2.4 Equipment Failure Rates............................................................................................... 10 2.5 System Reliability......................................................................................................... 12 2.6 Life Cycle Cost ............................................................................................................. 14 2.7 Inspection, Maintenance, and Replacement.................................................................. 16 2.8 Summary ....................................................................................................................... 18
3
STATE OF THE INDUSTRY .............................................................................................. 19 3.1 Historical Overview ...................................................................................................... 19 3.2 Present Situation in the U.S. ......................................................................................... 21 3.3 General Industry Trends ............................................................................................... 22 3.4 Substations .................................................................................................................... 23 3.5 Underground Cables ..................................................................................................... 26 3.6 Wooden Poles ............................................................................................................... 29 3.8 Summary ....................................................................................................................... 31
4
AGING INFRASTRUCTURE AT PG&E ........................................................................... 32 4.1 Overview of Aging Infrastructure at PG&E ................................................................. 32 4.2 Substations .................................................................................................................... 33 4.3 Underground Cables ..................................................................................................... 35 4.4 Wood Poles ................................................................................................................... 40 4.5 Back Tie Capacity......................................................................................................... 41 4.6 Summary ....................................................................................................................... 43
5
PG&E APPROACH TO ASSET MANAGEMENT ............................................................ 45 5.1 Overview of Asset Management................................................................................... 45 5.2 Industry Trends ............................................................................................................. 46 5.3 Asset Management at PG&E ........................................................................................ 47 5.4 Summary ....................................................................................................................... 51
6
CONCLUSIONS................................................................................................................... 52
APPENDIX A. Qualifications of Richard E. Brown.................................................................... 54
18-5
(PG&E-4)
1 INTRODUCTION KEMA has been retained by Pacific Gas and Electric Company (PG&E) to review its distribution aging infrastructure situation as it relates to the 2007 general rate case filing. Work for this report began in May 2005 and was completed in July 2005. This report was written by Richard E. Brown, a Senior Principal Consultant with KEMA. Qualifications of the author are provided in Appendix A. In writing this report, KEMA was asked by PG&E to provide an objective assessment of distribution aging infrastructure including (1) the present state of the industry, (2) the present state of PG&E, and (3) the ability of the PG&E asset management processes to make good financial decisions. Issues related to aging infrastructure are forward-looking by nature. Effective management of aging infrastructure requires the anticipation of future equipment deterioration, escalating risks, and escalating costs. There are many uncertainties when making these predictions, but failure to act based on this uncertainty is not prudent from the standpoints of safety, reliability, or cost minimization. This report is organized as follows. It first presents an overview of aging infrastructure with the intent of providing a foundation of knowledge upon which subsequent sections are built. It then discusses the state of the U.S. industry with respect to aging equipment on distribution systems. Next, the report assesses the aging distribution infrastructure at PG&E with a focus on substations, underground cables, wood poles, and back tie capacity. It then reviews the PG&E approach to asset management, and examines the ability of PG&E to make good financial decisions with regards to aging infrastructure. The report ends by summarizing its major findings and conclusions.
18-6
(PG&E-4)
2 OVERVIEW OF AGING INFRASTRUCTURE This section provides an overview of the basic concepts, terminology, and mechanisms surrounding aging infrastructure. The intent is to provide a foundation of knowledge upon which subsequent sections are built. Nothing in this section is intended to relate specifically to the situation at PG&E. 2.1
Equipment Aging
At face value, equipment aging seems to be a simple concept. A piece of equipment can be brand new, five years old, ten years old, fifty years old, and so forth. Unfortunately it is not as simple as this. It is possible for a piece of equipment to be purchased fifty years ago, installed forty five years ago, taken out of service thirty years ago, put back into service twenty five years ago, and had most of its critical parts replaced ten years ago. What age should be assigned to this piece of equipment? Before answering this question, it is beneficial to consider another question, “Why do we care about the age of equipment?” There are many answers, but the following are most important: Key considerations of equipment aging - The likelihood of failure tends to increase - Maintenance costs tend to increase - Replacement parts can become difficult to obtain - Old equipment may become technologically obsolete Although all of the above considerations are important, the first is the most difficult to address. This is because issues such as maintenance costs, spare parts availability, and technological obsolescence can be addressed in a business case. If it is cheaper to buy new equipment than maintain old equipment, the new equipment should be purchased. If the value proposition of new technology is compelling, old technology should be replaced with new technology. But what about equipment that is neither expensive to maintain nor obsolete? It is simply old and getting older. It should be emphasized that old equipment is not undesirable simply because it is old. It is undesirable because it may have a high likelihood of failure. In fact, any piece of equipment that has a high likelihood of failure is also undesirable. Utilities are not so much interested in equipment age. Rather, utilities are interested in the condition of the equipment, and often use age as an indicator of equipment condition. Typically the best measure of age when used for this purpose is based on purchase date or installation date. How should equipment aging be viewed? - Age can often be used as an indication of equipment condition - For this purpose, age should be based on purchase date or installation date Given that age is important, it is also important to understand the age characteristics of a population of equipment. This is typically done through equipment age profiles.
18-7
(PG&E-4)
2.2
Equipment Age Profiles
Consider a utility that has installed one transformer each year for the last fifty years. This utility will have one transformer that is one year in age, one that is two years in age, one that is three years in age, and so forth up until fifty years in age. If you are interested in the age characteristics of these transformers, you could look at a list of all the transformers and their corresponding ages, but this would be difficult to interpret. You could also use statistical measures like average age and median age, but these measures do not provide enough information. A better way to look at the data is to use equipment age profiles and survivor curves. An equipment age profile is simply a histogram showing the percentage of equipment within various age categories. For large populations, it is useful to have a separate age category for each year. For smaller populations, it may make sense to group equipment based on five-year age categories (1-5, 6-10, 11-15, etc.) or ten-year age categories (1-10, 11-20, 21-30, etc.). A sample age profile for underground cable is shown in Figure 2.1. The blue squares represent the percentage of cable that is each year in age. For instance, about 2.7% of all cable is one year in age, as shown by the left-most blue square. Similarly, about 3.5% of all cable is twenty years in age, as shown by the blue square directly above the twenty on the horizontal axis. Equipment age profiles contain a lot of data, but can sometimes be difficult to interpret since equipment purchases and installations can vary widely from year to year, resulting in an erratic histogram. For this reason, it is often times beneficial to look at a “survivor curve,” which represents the percentage of equipment that has survived up to any given age. In the case that 5% of all cable is between zero and one year in age, it follows that 95% of all cable has survived passed the age of one. Similarly, if another 5% of cable is between one and two years in age, it follows that 90% of all cable has survived passed the age of two. The calculation of the amount of cable that has survived passed all ages can be calculated in a similar manner, resulting in a survivor curve. The survivor curve corresponding to the histogram in Figure 2.1 is shown as a thick line in Figure 2.1. This curve shows that about 50% of all cable is older than 18 years in age. This is determined by following the 50% line on the right vertical axis over to where it crosses the thick line at a value of about 18 years. In a similar way, it can be seen that about 20% of all cable is older than 29 years and about 10% of cable is older than 33 years. Very little cable is older than 40 years in this case.
18-8
5.0%
100%
4.5%
90%
4.0%
80%
3.5%
70%
3.0%
60%
2.5%
50%
2.0%
40%
1.5%
30%
1.0%
20%
0.5%
10%
0.0%
% This age or older .
% Exactly This Age .
(PG&E-4)
0% 0
5
10
15
20
25
30
35
40
45
50
Age of Cable in Years Figure 2.1. The squares connected by the thin line represent the percentage of underground cable at each specific age. For example, the left-most square shows that 2.7% of all underground cable is between zero and one year old. The thick line is a survivor curve and shows the percentage of cable that is older than each age level. For example, the thick line crosses the point “Age 25” at a y-axis value of 30%. This means that 30% of all cable is older than 25 years in age.
The following sections of this report will use equipment age histograms and survivor curves similar to those shown in Figure 2.1.
2.3
Population Aging Behavior
Each class of equipment is constantly aging and being rejuvenated. Each year, the entire population of equipment becomes older. Each year, new equipment is installed which makes the average equipment age lower. Each year, some equipment will typically fail and be replaced. This will also make the average equipment age lower. Factors Affecting Average Population Age - Chronological aging of equipment - Installation of new equipment - Replacement of old equipment with newer equipment (often after a failure) Consider a population of 1000 new transformers. Initially, the average age of this population is 1 year. As long as there are no failures, the average age will increase each year. The second year will have an average age of 2 years, the third year will have an average age of 3 years, and so forth. Eventually, older equipment will start to fail and be replaced by new equipment. This will slow the rate of increase for the average population age. Eventually, the rate of rejuvenation due to replacement will fully offset the rate of chronological aging and the average equipment age will stabilize. However, this process may take decades and involve oscillations where average
18-9
(PG&E-4) population age rises and falls dramatically as large percentages of old equipment fail together and are replaced within a short period of time. Growth tends to mask the signs of population aging. Consider a population of 1000 new transformers with an average age of 30 years. In ten years, if no transformers are replaced, the average age of transformers will be 40 years. In the same ten years, if growth is about 7% per year, the number of transformers will typically double. Therefore, the final population will consist of the initial 1,000 transformers plus 1,000 new transformers. The initial transformers have an average age of 40 years, and the new transformers have an average age of 5 years. Therefore, the final population will have an average age of 22.5 years, which is lower than the average age of the original 1,000 transformers. In the situation described in the previous paragraph, it may not seem like there is an aging infrastructure problem since average equipment age is going down. However, there is a large population of equipment that is all getting old at the same time, which is a major concern. Therefore, it is typically better to view aging infrastructure by looking at the amount of very old equipment on the system rather than looking at average equipment age. Important Points Regarding Population Age - Average age is not as important as the amount of very old equipment - Low average age may be a result of high load growth. This does not eliminate the obligation of a utility to take care of customers in older parts of the system In the same way that high load growth can mask the effects of aging infrastructure, low load growth can have the opposite effect. In addition to making aging statistics more pronounced, a higher percentage of utility resources must focus on deteriorating equipment as opposed to new construction. This type of shift can be disruptive because spending related to new construction typically corresponds to increases in revenue. In contrast, efforts related to aging infrastructure typically increase spending without corresponding increases in revenue, at least without any associated rate adjustment.
2.4
Equipment Failure Rates
It might seem reasonable to conclude that new equipment fails less than old equipment. When dealing with complex components, this is usually not the case. In fact, newly installed electrical equipment has a relatively high failure rate due to the possibility that the equipment has manufacturing flaws, was damaged during shipping, was damaged during installation or was installed incorrectly. This period of high failure rate is referred to as the infant mortality period or the equipment break-in period. If a piece of equipment survives its break-in period, it is likely that there are no manufacturing defects, that the equipment is properly installed, and that the equipment is being used within its design specifications. It now enters a period referred to as its useful life, characterized by a nearly constant failure rate that can be accurately modeled by a single scalar number for that class of equipment.
18-10
(PG&E-4)
As the useful life of a piece of equipment comes to an end, the previously constant failure rate will start to increase as the components start to wear out. That is why this time is referred to as the wear-out period of the equipment. During the wear-out period, the failure rate of a component tends to increase exponentially. Upon failure, the component should be replaced. A graph that is commonly used to represent how a component’s failure rate changes with time is the bathtub curve. The bathtub curve begins with a high failure rate (infant mortality), lowers to a constant failure rate (useful life), and then increases again (wear-out). Another name for the bathtub curve is the bathtub hazard function. The use of the term “hazard rate” is common in the field of reliability assessment and is equivalent to the failure rate of the component. A more detailed curve used to represent a component’s hazard function is the sawtooth bathtub curve. Instead of using a constant failure rate in the useful life period, this curve uses an increasing failure rate. The increase is attributed to normal wear, and can be mitigated by periodic maintenance. This is analogous to changing the oil in an automobile. If done regularly, the reliability of the car will not degrade substantially. If changing the oil is neglected, reliability will quickly degrade and the probability of a failure occurring increases accordingly. If performing maintenance on the component reduces the failure rate to the same level each time, it is referred to as perfect maintenance. A bathtub function and a sawtooth bathtub function are shown in Figure 2.2.
1
Failure Rate (per year)
Failure Rate (per year)
In the real world, maintenance is rarely perfect. After each maintenance effort, component reliability will usually be a bit worse than the last time maintenance was performed. If performing maintenance on the component does not reduce the failure rate to the same level each time, it is referred to as imperfect maintenance. A further complication is that the failure rates after maintenance can often increase temporarily. This phenomenon, similar to infant mortality, is due to the possibility of maintenance crews causing damage, making errors during re-assembly, leaving tools inside equipment and so forth. If the maintained equipment survives for a short period of time, the maintenance was probably performed properly and failure rates are decreased accordingly.
0.8 Infant Mortality/Break-in Period
0.6
Wear Out
0.4 0.2 Useful Life 0 0
10
20
30
Com ponent Age (years)
1 0.8 Mean Failure Rate During Useful Life
0.6
Maintenance Reduces Failure Rate
0.4 0.2 0 0
10
20
Com ponent Age (years)
Figure 2.2. A standard bathtub curve (left) and a sawtooth bathtub curve (right). The standard bathtub curve is characteristic of the failure rates of many electrical components that are prone to shipping damage and installation errors. The standard bathtub curve is usually an approximation of the sawtooth bathtub curve, which models the increasing failure rate of a component between maintenance and shows a reliability improvement after maintenance has been performed.
18-11
30
(PG&E-4)
A standard bathtub curve is an approximation of a sawtooth bathtub curve. It models the useful life as the average useful life of the sawtooth curve. This approximation is sufficient for most reliability models, but a full sawtooth curve must be used if decisions about maintenance are to be made. Bathtub curves are commonly taught in reliability theory, and typically have been applied to equipment with moving parts. For electric utilities, it is fair to ask whether this type of failure curve accurately reflects electrical equipment aging. The short answer is yes, except that there is a gradual increase in failure rates during the useful life period, similar to what one might expect with imperfect maintenance (even though certain types of electrical equipment, like cables, are typically not maintained). For electrical equipment with insulation subject to thermal aging (such as transformers and cables), failure rates tend to increase exponentially with age. This means that after an initial infant mortality period, failure rates tend to be low during the design life of the equipment. After a certain point, failure rates will tend to dramatically rise as the electrical and mechanical strength of the insulation becomes insufficient to withstand physical or electrical disruptions on the system. 2.5
System Reliability
Equipment failures are not necessarily bad unless the failures are unsafe, expensive, or result in poor customer reliability. For example, some systems are designed to handle the failure of any single piece of equipment by having built-in redundancy. This redundancy is expensive, but events involving a single failure will not result in any customer interruptions. Most utilities track the reliability of their system using reliability indices, which are average values across the system. The most widely used reliability indices are averages that weight each customer equally. Customer-based indices are popular with regulating authorities since a small residential customer has just as much importance as a large industrial customer. They have limitations, but are generally considered good aggregate measures of reliability and are often used as reliability benchmarks and improvement targets. Formulae for customer-based indices include (unless otherwise specified, interruptions refer to sustained interruptions): System Average Interruption Frequency Index: Total Number of Customer Interruptions SAIFI = /yr Total Number of Customers Served System Average Interruption Duration Index: ∑ Customer Interruption Durations SAIDI = Total Number of Customers Served
18-12
hr/yr
(PG&E-4)
Customer Average Interruption Duration Index: ∑ Customer Interruption Durations hr CAIDI = Total Number of Customer Interruptions SAIFI is a measure of how many sustained interruptions an average customer will experience over the course of a year. For a fixed number of customers, increased equipment failures due to aging infrastructure will have an adverse impact on SAIFI. SAIDI is a measure of how many interruption hours an average customer will experience over the course of a year. For a fixed number of customers, SAIDI can be improved by reducing the number of interruptions or by reducing the duration of these interruptions. Like SAIFI, increased equipment failures due to aging infrastructure will have an adverse impact on SAIDI. CAIDI is a measure of how long an average interruption lasts, and is used as a measure of utility response time to system contingencies. CAIDI can be improved by reducing the length of interruptions, but can also be reduced by increasing the number of short interruptions. Therefore, aging infrastructure could either make CAIDI worse or better. If the increased failures due to aging equipment are faster on average to fix, CAIDI will improve. If the increased failures due to aging equipment are slower on average to fix, CAIDI will become worse. Typically most distribution systems are radial in nature. This means that a single failure will always result in customer interruptions. This type of single failure is sometimes referred to as a “N-1 contingency” since the original system consisting of “N” components must now make do with only “N-1” components. This type of system will, for the most part, experience twice as many customer interruptions if all failure rates double. Some systems are designed to handle any N-1 contingency, and are called “N-1 secure.” These systems are typically much more reliable than radial systems since multiple failures must occur for customer interruptions to occur. For these systems, it may seem natural to dismiss the effects of aging infrastructure since single failures do not result in customer interruptions. In fact aging infrastructure has a much more dramatic impact on redundant systems than with radial systems. This is because the probability of two overlapping outages occurring increases with the square of equipment failure rates. If failure rates increase by a factor of five on an N-1 secure system, customer interruptions will increase by a factor of 25. This calculation optimistically assumes that failures are independent. In reality, the situation will probably be worse, since failures will increase loading on redundant equipment, increasing the probability of failure when compared to normal operations.
Important Points Regarding System Reliability - For radial systems, interruptions will tend to increase linearly with failure rate - For N-1 redundant systems, interruptions will tend to increase with the square of failure rate. This implies that aging infrastructure will have the greatest impact on redundant systems. Most state regulatory commissions now have some type of reliability regulation related to reliability indices such as SAIFI and SAIDI. In almost all cases, the expectation is for reliability per18-13
(PG&E-4) formance to improve over time. This will become increasingly more difficult as certain types of aging infrastructure begin to show increases in failure rate, resulting in increases in costs and increases in customer interruptions.
2.6
Life Cycle Cost
The regulatory mandate of an electric utility is to provide adequate levels of service to all customers in its service territory in a manner that will result in the lowest possible rates for customers. Based on estimates of these costs and the need for a utility to attract capital investment, regulators will then set rates that allow utilities to recover costs and make a reasonable profit. Cost typically refers to cash payment obligations related to business expenses such as products, services, labor, real estate, insurance, interest, and taxes. It is important to accurately capture all costs when examining spending alternatives, and to understand the concepts of time value of money, sunk costs, avoided costs and opportunity costs.
Sunk Cost – a cost that has already been incurred. From an economic perspective, sunk costs should not impact future decisions. In reality, sunk costs often influence later decisions due to non-economic issues such as attempts to justify these prior expenditures. Avoided Cost – a cost that would have normally been incurred, but is not incurred as a result of a decision. Avoided costs are equivalent in value to cash costs and should be included in all economic decisions. Opportunity Cost – the cost of the next best economic decision. In business, opportunity cost is typically considered to be the cost of buying back bank loans, bonds or stock. The weighted expected return of these cash sources is referred to as the weighted average cost of capital (WACC), and is the minimum opportunity cost hurdle rate of a company. Costs can also be divided into initial costs and recurring costs (sometimes referred to as one-time costs and annual costs). Initial costs are typically incurred during the procurement and construction of a capital project, and do not repeat once they are paid. Recurring costs are the carrying charge of an asset and must be periodically paid as long as the asset is still owned. Table 2.1 provides a list of commonly considered initial and recurring costs. There may also be costs associated with the disposal and removal of equipment at the end of its life. Table 2.1. Common initial and recurring costs. Initial Costs Procurement Cost Material Cost Shipping Cost Labor Cost Equipment Rental Cost Commissioning Cost
Recurring Costs Operating Cost Maintenance Cost Income Tax Property Tax Insurance Premiums (Depreciation)
18-14
(PG&E-4)
Depreciation of an asset on a company’s general ledger is treated as a negative recurring cost since it reduces a company’s tax burden. Since depreciation lowers earnings by an equal amount, the negative recurring cost is computed by multiplying the amount of depreciation by the company’s marginal tax rate. The main difficulty in comparing the cost of different projects occurs when initial and recurring costs are different and/or occur at different points in time. From both a rate-making perspective and an investment perspective, the best method of combining such costs into a single measure is referred to as life cycle costing, typically measured by Net Present Value (NPV). Present value is a method of measuring and comparing costs that occur at different points in time. For example, a utility might value being paid $100 today just as much as it would value being paid $110 in one year. In this case, $110 in one year is said to have a present value of $100, and $100 is said to have a future value of $110 in one year. Present value is based on the fundamental economic principle that money received today is more valuable than money received tomorrow, and is calculated based on a discount rate, d. This is the perceived rate of depreciation in one year. For example, if $115 paid in one year is worth only $100 if it is paid today, the discount rate is said to be 15% per year. Present value for a cost, C, at year t can then be computed as follows:
Present Value = C ⋅ (1 + d ) − t The net present value of a piece of equipment is computed by summing the present values of all individual cash transactions. This can include both costs, which have a negative present value, and revenues, which have a positive present value. m
NPV = ∑ Ri (1 + d ) i =1
NPV R C tr,i tc,j
= = = = =
−t r , i
n
− ∑ C j (1 + d )
−tc , j
j =1
Net Present Value Revenue Cost Year that revenue i is realized Year that cost i is incurred
Net present value is a powerful economic metric that is commonly used in reliability and other engineering analysis. It is of further interest because it is the method typically used by financial analysts to compute the intrinsic value of companies. The value of a company is equal to the net present value of future cash flows discounted by the weighted average cost of capital. This value minus the value of long-term debt obligations is equal to the value of outstanding stock. The value of outstanding stock divided by the number of outstanding shares should be equal to stock price.
18-15
(PG&E-4) Life cycle costing is critical when addressing aging infrastructure. This is true since spending decisions related to aging infrastructure can be deferred, which has value since money today is worth more than money in the future. In addition, aging infrastructure decisions often require a trade-off between recurring expenses and one-time expenses. These trade-off can only be quantified by considering the time value of money.
Key Points Regarding Life Cycle Cost - Life cycle costing is required in order to make good decisions relating to aging infrastructure - Net Present Value (NPV) is the correct measure to use for life-cycle costing. NPV can also be converted to an annualized recurring cost, which is equivalent. - For aging infrastructure decisions, NPV should be computed using a discount factor equal to the Weighted Average Cost of Capital (WACC). In certain situations, it may be appropriate to risk-adjust adjust this discount factor.
2.7
Inspection, Maintenance, and Replacement
Many types of distribution equipment require inspection, testing and/or maintenance to ensure proper operation and minimize the probability of failures. Maintenance strategies can be broadly categorized as run-to-failure, periodic, condition based and reliability centered. Run-to-failure is the simplest maintenance strategy. After installation, equipment is not inspected or maintained until a failure occurs. This is cost effective for non-critical components with minimal maintenance requirements. It is also commonly used for equipment that is difficult or expensive to obtain meaningful condition data. Periodic maintenance is the next simplest maintenance strategy. At specific time intervals, certain maintenance procedures are performed on equipment regardless of equipment condition or equipment criticality. Intervals are usually time-based, but other measures such as number of operations are also used. This has historically been the most common form of maintenance for major pieces of equipment, with periodic inspection and maintenance activities often being performed according to manufacturer recommendations. However, many types of equipment should not be maintained until a critical level of deterioration has occurred. This is because the very act of maintenance can cause damage to otherwise healthy equipment (i.e., don’t fix it if it isn’t broken). Many utilities recognize this and are moving away from periodic maintenance and towards condition-based maintenance. Condition-based maintenance (CBM) monitors equipment and only requires maintenance when signs of deterioration become evident. Condition assessment techniques are too numerous to list exhaustively, but include techniques such as visual inspection, acoustical inspection, infrared thermography, voltage withstand testing, partial discharge testing, laboratory testing, and many other techniques. Pure CBM will perform a basic inspection on a piece of equipment and perform additional testing and maintenance as a result of this inspection. In addition, the future date and extent of the next inspection is based on the present condition.
18-16
(PG&E-4) There is always a danger that a piece of equipment will fail between inspections, especially if the inspections occur far apart in time. For this reason, there is an increasing trend towards the use of real-time condition monitoring for critical pieces of equipment such as large substation transformers. These devices will monitor key indicators of condition and report and signs of incipient failure to dispatchers over the SCADA system (a real-time communications system). Although far from perfect, such systems allow utilities to move closer to “just-in-time” maintenance and replacement. Reliability-centered maintenance (RCM) is a process used to determine maintenance requirements based on equipment condition, equipment criticality and cost. Effective applications of RCM result in maintenance schedules that maximize system reliability by maintaining components that are most likely to fail, have a high impact to customers when they fail and can be maintained for a reasonable cost. This holistic approach makes intuitive sense, but is difficult to implement. RCM has been successfully applied to power plants, is beginning to be applied to substations, and is in its infancy for feeder maintenance and vegetation management. When determining the optimal maintenance strategy for a specific piece of major equipment, it is often possible to justify replacement based on escalating failure costs. For example, consider an old substation transformer with an expected remaining life of five years. Due to its deteriorated condition, this transformer has to be frequently inspected and maintained. The NPV of these inspection and maintenance costs can be computed and compared to the cost of replacement. In this case, the appropriate comparison is to five years worth of interest on the replacement cost of the transformer. The transformer will have to be replaced in any case, but the money spent on replacement today could be put in the bank earning interest until replacement in the future. If the NPV of five years worth of inspection and maintenance costs is more than five years worth of interest on the cost of a new transformer, the transformer should be replaced today. Of course, it is impossible to know with certainty when a piece of equipment will fail in the future. Many times good condition data is simply not available. Even if it is, it is difficult to convert this condition data into failure probabilities, which are required for a NPV analysis. For many engineers within utilities, there is an aversion to replacing equipment when it may have remaining life, partly due to this uncertainty. However, this aversion is in large part unwarranted. Equipment inspection, maintenance, and replacement decisions are not made in isolation. Many thousands of decisions are typically made each month. These decisions represent a portfolio that has an inherent performance in terms of cost, reliability, and risk. If professional investors purchase an individual stock, they are not surprised or disappointed if the value of the stock declines. Instead, these investors purchase a portfolio of stocks, and are only interested in the performance of the portfolio. In a similar manner, many inspection, maintenance, and replacement decisions will end up being sub-optimal due to a lack of perfect knowledge. This is perfectly natural as long as the overall portfolio of inspection, maintenance, and replacement decisions has good performance. It is always justifiable to spend money now to save money later, as is sometimes the case when replacing deteriorated but functioning equipment. However, there are also performance and risk issues associated with equipment failures. Even if it is the cheapest option to never trim trees,
18-17
(PG&E-4) doing so may result in unacceptable reliability to customers. Regulations require that utilities provide adequate levels of service, and utilities must therefore spend money to achieve these levels of service even if life cycle cost is not minimized. Stated differently, utilities must strive to provide adequate levels of service for the minimum life cycle cost. Equipment spending and system performance should ideally be considered simultaneously and rigorously. In practice this is difficult due to data limitations and a lack of analytical tools. The last difficulty with equipment inspection, maintenance, and replacement decisions relates to risk. In addition to technical risk, this might include other aspects of risk including social and political. There may be many “headline events” such as interruptions to major public facilities that may not contribute substantially to overall customer interruptions. However, this is a high profile situation that may result in an adverse response from city officials and, ultimately, strained regulatory relations. The costs associated with this outcome are real and substantial, but difficult to quantify. Therefore it is difficult to decide how much to spend in consideration of these types of risk. In many cases, qualitative risk factors play an important role in spending decisions. This situation is not necessarily inappropriate, but makes it difficult to prioritize spending and to know how much risk-based spending is justifiable.
Key Points Regarding Inspection, Maintenance, and Replacement - The basic ways to inspect and maintain equipment are run-to-failure, periodic maintenance, condition-based maintenance (CBM), and reliability-centered maintenance (RCM). - Sometimes it is less expensive to replace deteriorated equipment today rather than incur high maintenance costs. - Sometimes it is justified to spend more money in inspection and maintenance in order to ensure adequate service to customers. - Sometimes it is justified to spend more money in inspection and maintenance in order to mitigate unacceptable risks.
2.8
Summary
This section has provided an overview of the basic concepts, terminology, and mechanisms surrounding aging infrastructure. This includes the basics of equipment aging; representation of equipment age profiles; population aging behavior; equipment failure rate characteristics; system reliability life-cycle costing and net present value; and inspection, maintenance, and replacement decisions. Nothing in this section is intended to relate specifically to the situation at PG&E, but careful reading will magnify the clarity and utility of the rest of this report.
18-18
(PG&E-4)
3 STATE OF THE INDUSTRY This section provides a general overview of the state of the electric utility industry with an emphasis on distribution infrastructure in the United States. The approach is to discuss the history of the industry and the impact of this history on equipment age, equipment condition, and equipment investment trends. Also included are discussions on several classes of assets that are of particular interest to PG&E with regards to aging infrastructure. The intent is not to make any particular value judgments with regards to PG&E (this will be done in Section 4 and Section 5). Rather, it is to compare the asset age profiles of PG&E with the asset age profiles of peer utilities, and to examine what utilities facing similar situations are pursuing with regards to failure mitigation and increases in investment.
3.1
Historical Overview
The electric power industry began in the late 1800s as a component of the electric lighting industry. At this time, lighting was the only application for electricity, and homes had other methods of illumination if the electricity supply was interrupted. Electricity was essentially a luxury item and reliability was not an issue. As electricity became more common, new applications began to appear. Examples include electric motors, electric heating, irons, and phonographs. People began to grow accustomed to these new electric appliances, and their need for reliable electricity increased. This trend culminated with the invention of the radio. No non-electrical appliance could perform the same function as a radio. If a person wanted to listen to the airwaves, electricity was required. As radio sales exploded in the 1920s, people found that reliable electricity was a necessity. By the late 1930s, electricity was regarded as a basic utility. From rough1y 1935 to the early 1970s, distribution investment was massive, generally between 7% and 15% per year as a percentage of electricity revenues (see Figure 3.1). This has several implications. First, overall utility growth was consistent and predictable. This led to an attractive combination of increasing revenue, increasing efficiencies of scale, and generally decreasing rates. In this situation everybody is generally happy, and the primary focus of utilities is on system capacity expansion. In addition, the massive distribution investment from 1935 to 1975 resulted in a doubling of installed distribution equipment every five to ten years. Essentially the distribution systems were self-rejuvenating, with the amount of old equipment always being small as a percentage of total equipment population.
18-19
(PG&E-4)
60%
Investor Owned Utility Capital Invested as % of Electricity Revenues (Source: EPRI) 50%
40%
EPRI Forecast 30%
Generation 20%
Transmission 10%
Distribution 19 25 19 30 19 35 19 40 19 45 19 50 19 55 19 60 19 65 19 70 19 75 19 80 19 85 19 90 19 95 20 00 20 05 20 10 20 15 20 20
0%
Figure 3.1. Utilities invested heavily into their distribution systems up until the mid-1970s. At this point, the growth of electricity demand slowed, and infrastructure investment has a corresponding decline. For the first time, distribution systems were aging overall. Further, the large amount of equipment installed before the mid-1970s is now all over the age thirty years.
The industry reached an inflection point with the oil crisis of the early 1970s. Load growth declined from over 7% to less than 2%. In some areas, energy conservation efforts actually resulted in a decline in electricity consumption. For the first time in more than forty years, utilities were faced with poor growth prospects, both in terms of revenue and infrastructure expansion. The effect of these pressures was a gradual decrease in distribution capital spending (as a percentage of utility revenue) that continues up to the present day. In the middle of the 1980s, many utilities began to complete major capital projects that were initiated in the 1970s in the anticipation of robust load growth (primarily power plants). When this growth did not materialize, many available resources could not be put to work on new construction. At the same time, many utilities had high allowed rates-of-return due to the high interest rates of the 1970s. This combination led many investor-owned utilities to self-imposed rate freezes and aggressive cost reduction programs. In many ways, the electric utility trends starting in the 1980s can be summarized as a shift from an engineering-focused industry towards a business-focused industry. Prior to the 1980s, utilities executives commonly had engineering backgrounds. Since the 1980s, utility executives are more likely to have business or legal backgrounds. The end results tended towards employee downsizing, reductions in spending, operational efficiency initiatives.
18-20
(PG&E-4)
Industry Trends in the 1980s - A massive reduction in personnel, especially with regards to higher-paid technical employees who best understand the technical issues related to aging equipment and system issues. - Efforts to reduce capital spending by loading existing equipment to higher levels than historically were allowed. This higher loading accelerates condition deterioration, and may limit the ability of dispatchers to restore customers after a fault occurs through system reconfiguration. - Efforts to reduce maintenance spending by extending maintenance cycles and adopting condition-based maintenance for certain classes of equipment. - Efforts to improve operational efficiency through the use of sophisticated software tools (financial, project management, maintenance management, etc.), performance management, outsourcing, benchmarking, and so forth. The cost-cutting trends of the 1980s were reinforced by the passage of the National Electric Policy Act in 1992 (NEPA). NEPA indicated that the industry was headed towards “de-regulation” and competition. During this time, each state also took its own approaches with regards to electric distribution regulation. Results varied by state, but the general effect was a re-doubled emphasis on cost cutting and efficiency improvement initiatives in order to maximize the chances of survival in this new environment. Nearly every aspect of utility efficiency improvement efforts since the 1980s has had a deleterious effect of aging infrastructure and the ability of utilities to address aging infrastructure. Much of the distribution equipment installed in the 1960s and early 1970s is reaching the end of its expected life. Due to NEPA, this equipment has been pushed harder and has been maintained less in the last decade. Many utilities have achieved substantial improvements in business performance, but are now seeing signs of increasing equipment failure rates. Many of these same utilities are finding it difficult to address this difficult issue due to a lack of senior technical staff and a heavily loaded workforce.
3.2
Present Situation in the U.S.
Today, most large investor-owned utilities have either begun to increase spending on aging infrastructure, have announced intentions to increase spending on aging infrastructure, or are beginning to experience signs of increased failure and are deciding how to address the problem. Since this is a direct result of electricity growth patterns since the 1960s, the situation is similar for many utilities, and is most problematic for utilities serving older cities. Perhaps the first signs of systematic aging infrastructure problems were witnessed in the summer of 1999, when the country experienced major electricity outages in New York City, Long Island, New Jersey, the Delmarva Peninsula, and Chicago. These events were investigated by the U.S. Department of Energy through a Power Outage Study Team (POST). The resulting POST report concludes the following (Page S-2):
18-21
(PG&E-4)
The POST investigations found that the aging infrastructure and increased demand for power have strained many transmission and distribution systems to the point of interrupting service. In the case of Commonwealth Edison in Chicago (now Exelon Corporation), these outages resulted in immediate and large increases in capital spending – more than one billion dollars from 1999 to 2004. When aging infrastructure began to manifest itself in 1999, many utilities were under rate freezes, and had a limited ability to fund large increases in capital spending in the short term. However, the pressure to proactively address the aging infrastructure problem resulted in a substantial increase in rate case activity. From 1996 to 2000, the average amount requested was $742 million. From 2001 through 2004, the average amount requested per year rose to more than $2 billion (see Table 3.1). Table 3.1. Number of rate case filings and aggregate amount requested from 1995 through 2004. As of December 1
Number of Rate Case Filings
Aggregate Amount Requested (MUSD)
Electric
Gas
Electric
Gas
2004
21
21
2,171.8
661.8
2003
20
23
2,400.3
970.5
2002
16
20
1,692.7
351.4
2001
21
12
1,816.8
560.4
2000
12
8
1,003.1
161.7
1999
10
13
759.0
562.9
1998
14
8
894.8
546.1
1997
13
11
624.9
161.9
1996
16
18
429.8
266.7
This situation in California is different when compared to many other states since regulated utilities in California are required to file for rate cases on a 3-year rate case cycle. In other states, some utilities are subject to rate freezes lasting more than a decade. From this perspective, it is important to understand that utilities in many other states have started to file new rate cases, in large part to address aging infrastructure issues, after long periods without base rate adjustments.
3.3
General Industry Trends
Although the specific approaches to aging infrastructure are different for each utility, there are some general trends in the industry. These trends generally relate to increased condition monitoring, increased efforts at life extension, more rigorous approaches to replacement decisions, and efforts to optimize spending in a rigorous manner. Not all utilities are making progress in these areas, but best-in-class performers are effectively utilizing these approaches to help address aging infrastructure. A brief summary of industry trends in each of these areas is now provided.
18-22
(PG&E-4)
Condition monitoring. Most medium and large utilities are placing more emphasis on condition monitoring, and are increasingly using condition monitoring data to help make operational and maintenance decisions. Often these efforts focus on more aggressive inspection and testing. Techniques that are becoming more popular include infrared inspections, dissolved gas-in-oil analysis, frequency response signatures, and many others. On-line condition monitoring is also becoming more popular, especially for substation equipment that can be cost-effectively monitored through the SCADA systems. On-line techniques range from simple alarms (e.g., temperature, pressure) to continuous monitoring of dissolved gases in oil. Some of the newest equipment can be equipped by the vendor with self-diagnostic capability that will interpret monitoring data and automatically notify operators if significant equipment deterioration or incipient failure has been detected. Equipment Life Extension. Equipment replacement is often expensive, resource intensive, and operationally disruptive. For this reason, utilities are increasingly looking at life extension strategies as a way to defer replacement. Often times an equipment life extension program is coupled with a condition monitoring program. When the condition of a piece of equipment reaches a certain degree of deterioration, life extension options are examined for economic attractiveness. An example is wooden pole programs that require periodic inspections. If a pole is likely to fail before the next inspection, the pole is treated, reinforced, or replaced depending upon the nature of deterioration. Similar strategies can be applied to transformers, circuit breakers, underground cables, and many other types of equipment. Repair versus Replacement Decisions. Utility practices still vary widely in this area, with some utilities making explicit repair-versus-replace decisions, while others will only replace equipment if it fails or becomes overloaded. In general, only the largest utilities in the U.S. are moving towards specific replacement criteria for major equipment. Most utilities are basing replacement decision on a combination of condition data and qualitative judgment. Economic Optimization Strategies. Optimal, optimize and optimization are terms prone to misuse. In its pure form optimization tries to minimize an objective function without violating any constraints (e.g. minimize life-cycle costs while meeting all reliability targets). In terms of aging infrastructure, few companies are making rigorous decisions, fewer are making purely economic decisions, and none that the author is aware of are employing strict optimization approaches. The rigor of optimization is simply not appropriate given the lack of data and uncertainty in assumptions. Pragmatic utilities are moving down the learning curve by gradually incorporating economic analysis into aging infrastructure assessment, and moving toward a more rigorous process that incorporates factors such as cost-to-benefit ratios and risk assessments.
3.4
Substations
Substations are where the distribution system begins. They receive power from one or more transmission lines, step voltage down to primary distribution levels, and supply a circuit breaker at the beginning of each feeder. From a distribution aging infrastructure perspective, the most critical pieces of equipment are substation power transformers and distribution circuit breakers, with most aging infrastructure problems occurring in old substations located in older cities.
18-23
(PG&E-4)
Several survivor curves for substation power transformers are shown in Figure 3.2, and similar curves for distribution circuit breakers are shown in Figure 3.3.These curves were assembled through a combination of public data and KEMA surveys, and show how the age of PG&E substation equipment compares to other large investor-owned utilities in the U.S. Figure 3.2 shows the survivor curves of substation transformers for eight large U.S. utilities, including PG&E (the PG&E curve is based on data supplied for the KEMA report “2004 Reliability Performance Assessment of San Francisco,” and therefore only includes data from the San Francisco operating division). All but one utility have at least 50% of their transformers older than 30 years in age. Three out of the eight utilities have more than 30% of their transformers older than 50 years in age. PG&E has the oldest population of transformers in this set, with more than half of all transformers older than 50 years in age. There are a lot of old transformers, but lightly loaded transformers can last many years. However, utilities have generally been loading transformers to higher levels in the past 15 years, and the combination of old chronological age and increased thermal aging has created significant deterioration in many transformers. Mainly because of this, substations were the first concern with regards to aging infrastructure, with emphasis starting after the national reliability problems in the summer of 1999. Many utilities have done a good job of assessing the condition of their older transformers and increasing condition monitoring for those at risk, but proactive replacement efforts have been generally modest and the overall population of transformers continues to get older. 100%
1. PG&E
90%
2. BG&E
% this age or older .
80%
3. OG&E
4
4. Cinergy
70% 1
5. SDG&E
60% 2
6. Utility A
5
50%
7. Utility B 8. Utility C
6
40% 30%
7
8
20% 10% 0% 0
10
20
30
40
50
60
70
80
Age of Substation Transformer
Figure 3.2. This figure shows the survivor curves of substation transformers for eight large U.S. utilities. All but one utility have at least 50% of their transformers older than 30 years in age, and PG&E has significantly older transformers than all other utilities shown in this graph.
18-24
(PG&E-4)
100%
1. PG&E
90%
2
2. SDG&E
% this age or older .
80%
3. Utility A 4. Utility B
70% 1
5. Utility C
60% 50% 40% 3 30%
5 4
20% 10% 0% 0
10
20
30
40
50
60
70
80
Age of Substation Circuit Breakers
Figure 3.3. This figure shows the survivor curves of substation circuit breakers (distribution class) for five large U.S. utilities. There is wide variation in this graph, indicating that some utilities have already replaced many of their older circuit breakers. For example, one utility only has 2% of circuit breakers older than 50 year, while others have corresponding values of 8%, 10%, 30%, and 37%. PG&E has the highest percentage of very old circuit breakers (older than 60 years in age).
At this point, it is important to remember that having an old population of equipment is not necessarily good or bad. In fact, good life-extension practices and condition-based maintenance intentionally strive to allow equipment to become older. However, older equipment in general will be more costly to maintain and more likely to fail than younger equipment. Figure 3.3 shows the survivor curves of substation circuit breakers (distribution class) for five large U.S. utilities. There is wide variation in this graph, primarily because some utilities have already replaced many of their old breakers. Often times this is due to a lack of spare part availability. If an old circuit breaker fails, many utilities will replace the unit rather than custommachine replacement parts. Overall, San Diego Gas and Electric has the oldest population of circuit breakers. However, PG&E has the highest percentage of very old circuit breakers (older than 60 years in age). As for the substation transformers, the PG&E curve is based on data supplied for the KEMA report “2004 Reliability Performance Assessment of San Francisco,” and therefore only includes data from the San Francisco operating division. From an industry perspective, there are a lot of old circuit breakers, but the population is somewhat younger when compared to substation transformers. From a criticality perspective, distribution circuit breakers are of special concern. This is because they are often the beginning of the radial distribution system. If a substation transformer fails, other transformers in the station can generally serve the load entire load with only a short inter18-25
(PG&E-4) ruption in service (often times with no interruptions at all). If a feeder circuit breaker experiences an internal fault, the entire feeder is interrupted. Worse, if a feeder circuit breaker is supposed to interrupt a downstream fault and fails to do so due to its deteriorated condition, the bus associated with the feeder will have to be de-energized. When this happens, all feeders served by this bus will experience a complete outage.
3.5
Underground Cables
Aging underground cable and the resulting increase in cable failures is one of the most important issues for all utilities in the U.S. that have significant underground distribution systems. However, the situation is not easy to characterize since underground cables can have many different characteristics that impact aging and failures. It is beyond the scope of this document to provide a comprehensive overview of underground cable systems, but a brief description of key features is now provided.
Insulation Type. Cable technology in the early 1900s typically wrapped conductors in oilimpregnated cable. Sometimes these cables are placed in oil-filled pipes, and are called pipe-type cable. Sometimes these cables are covered with a lead coating, and are called paper-insulatedlead-covered cable (PILC). Later, extrusion technologies allowed conductors to be covered with various typed of plastic insulation. Early technologies include polyethylene (PE), high molecular weight polyethylene (HMWPE). Newer technologies include cross-linked polyethylene (XLPE), tree-retardant cross-linked polyethylene (TR-XLPE), and ethylene-propylene rubber (EPR). Voltage Class. The thickness of insulation determines how much voltage the cable can withstand. Common voltage classifications for cables are 5 kilovolts, 15 kilovolts, 25 kilovolts, and 35 kilovolts. Sometimes, utilities will over-specify the insulation class to improve cable reliability (e.g., using 25-kV cable on a 15-kV system). Number of Conductors. Some cables have a single conductor, while other cables will combine three conductors or four conductors into a single cable. Concentric Neutral. Some cables have neutral wire strands distributed around the outside periphery of the insulation. Early versions of this sort did not protect the concentric neutral with a waterproof jacket. As such, the neutral strands tended to corrode and cause reliability problems. Waterproof jackets have largely fixed this problem.
18-26
(PG&E-4)
Jacket Insulation
Phase Conductor Concentric Neutral
Figure 3.4. Typical cable cross sections. Simple cables consist of a single insulated conductor (far left). Adding a concentric conductor allows the cable to serve single-phase loads without requiring a separate neutral cable (second to left). Since bare concentric neutrals can be subject to undesirable corrosion, they are often covered by a waterproof jacket (middle). If three-phase service is required, three single-phase conductors can be used. Alternatively, a single cable with three fully-insulated phase conductors can be used. These multi-phase cables are available both with and without a neutral conductor (second to right and right, respectively).
Installation. Most cable in urban areas is installed in concrete-encased ductbanks. These ductbanks provide mechanical protection and allow for cable to be replaced without excavation. In suburban areas, a large amount of cable is directly buried in the ground. This technique is cost effective, but makes cable replacement more difficult. Some cable is also installed in conduit that is not encased in a ductbank. This type of installation is called cable-in-conduit (CIC). Old PILC. Few utilities choose to install new PILC cable if it is possible to use other insulation types. Consequently, most existing PILC in the ground is old, commonly older than 50 years in age. Much of this cable has a history of reliable operation, and PILC cable that has been lightly loaded over its lifetime can be in very good condition at a very old age. However, utilities have tended to increase the loading of PILC cables over the last decade or more, and many are experiencing dramatic increases in PILC failures. Treeing. A major reliability concern pertaining to aging underground cables is electrochemical and water treeing. Treeing occurs when moisture penetration in the presence of an electric field reduces the dielectric strength of cable insulation. When moisture invades extruded dielectrics such as cross-linked polyethylene (XLPE) or ethylene-propylene rubber (EPR), breakdown patterns resembling a tree reduce the voltage withstand capability of the cable. When insulation strength is degraded sufficiently, voltage transients caused by lightning or switching can result in dielectric breakdown. The severity of treeing is strongly correlated with thermal age since moisture absorption occurs more rapidly at high temperature. Water treeing has been a widespread and costly problem for utilities with aging XLPE cable. (EPR cables have generally not encountered these problems). Figure 3.5 shows the cable age distributions for five utilities with extensive underground distribution systems. Since urban areas are typically served by a large percentage of underground distribution systems, it is common for utilities serving large cities to have the oldest populations of cable. In Figure 3.5, this is seen by the high percentage of old cable for Commonwealth Edison (serving Chicago), and San Diego Gas & Electric (serving San Diego). PG&E has a typical age distribution for its cables, but it is difficult to make specific aging cable judgments based on these curves since the problem is complex and can consist of many individual issues.
18-27
(PG&E-4)
KEMA has performed a cable survey with responses from twenty large U.S. utilities. Of these, nearly 70% have formal cable replacement programs that pro-actively replace old cable that is expected to be prone to failure in the future. The most common criterion for replacement is simply insulation type. For example, if a utility is experiencing a large number of failures on XLPE cable installed in the early 1980s, all cable of this type may be targeted for replacement. The second most common method is for utilities to track specific failures on each cable section. If multiple failures occur on a section, the section is replaced. Cable rejuvenation programs are also becoming more common. About 25% of surveyed utilities have cable rejuvenation programs that inject a gel into cables to restore deteriorated insulation properties. This technology allows the condition of direct buried cable to be improved without any excavation activity. Aging cables are a real problem for many utilities since they can adversely affect customer reliability and they are costly to replace. A few utilities have aggressively addressed this problem and, as a result, have already succeeded in replacing most of their first-generation extruded cables. In fact, about 60% of surveyed utilities feel that their aging cable problems will improve in the near future rather than get worse. The remainder feel that their present rate of cable replacement is not keeping up with the rate of cable deterioration due to aging.
100%
1. PG&E
90%
2. OG&E
4
% this age or older .
80%
3. ComEd
70%
4. SDG&E
60%
5. Xcel
50% 2
40% 30%
5
3
20% 1
10% 0% 0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
Age of Underground Cable Figure 3.5. This figure shows the survivor curves of underground cable (distribution class) for five large U.S. utilities. These graphs show that utilities primarily serving primarily older cities tend to have very old cable (ComEd and SDG&E). However this type of cable (paper insulation) can often have better reliability characteristics at a very old age than certain types of cable that are much younger (e.g., first-generation polyethylene insulation).
18-28
(PG&E-4)
3.6
Wooden Poles
Poles support overhead distribution equipment and are an important part of all overhead systems. Most poles are made of treated wood, but concrete and steel are also used. Typical distribution pole constructions are shown in Figure 3.6 (these examples are not exhaustive).
Figure 3.6. Common types of distribution pole construction.
18-29
Compact Dual Circuit Construction
Compact Armless Construction
Compact Construction with Post Insulators
Vertical Construction with Post Insulators
Symmetrical Construction
Crossarm Construction 2
Crossarm Construction 1
Single Phase Construction
Wood poles represent some of the oldest assets for many utilities; as such, wooden poles have become a potential aging infrastructure concern. However, the situation can be very different when comparing two utilities. Consider a utility in Florida. Every ten years or so a major hurricane is likely to occur. Since wooden poles are not designed to withstand hurricane-force winds, many of the wooden poles will be damaged and require replacement. In this situation, it is unlikely that a wooden pole will ever reach a point where age-related deterioration is a concern.
(PG&E-4)
100%
1. PG&E
90%
% this age or older .
1
2. OG&E
4
80%
3. Cinergy
70%
4. SDG&E 5. MidAmerican
60%
6. Xcel
50% 6 40% 2
30%
3
20%
5
10% 0% 0
10
20
30
40
50
60
70
80
90
100
110
Age of Wooden Pole
Figure 3.7. This figure shows the survivor curves of wooden poles (distribution class) for six large U.S. utilities. The age distributions are more similar that previous curves. For poles older than fifty years, results generally range from 9% to 20%, with only SDG&E scoring higher at 30%.
The situation is different in areas with mild weather. Utilities in these climates can easily have poles last seventy years and more. Figure 3.7 shows the age distribution of wooden poles for six large U.S. utilities. All of these utilities have between 5% and 15% of their poles greater than sixty years in age. The PG&E curve shows the highest number of extremely old poles (i.e., older than 80 years in age), reflecting the old nature of parts of the PG&E system and the mild climate of the PG&E service territory. Older poles are susceptible to rot and decay, but periodic pole inspections typically can identify such problems and determine whether problem poles should be treated, reinforced, or replaced. A typical inspection cycle is ten years, with ten percent of poles being inspected every year. Problems with aging poles have not yet manifested themselves. In part this is because of aggressive inspection and treatment programs. It is also because wooden poles under the right circumstances can last a long time. However, many utilities with an old population of poles are only replacing 0.5% of poles or less each year. For this amount to be sustainable, average pole life must be at least 200 years or more. Very few people believe this to be the case, and utilities will eventually have to increase the rate of pole replacement, even if this is strictly due to an increase in pole failures.
18-30
(PG&E-4)
3.7
Summary
For most U.S. utilities, distribution equipment is getting older both in terms of average age and the percentage of very old equipment. This situation is a direct result of the business environment. The oil crisis of the early 1970s resulted in low load growth, drastically reducing the amount of new equipment being installed. The National Electric Policy Act of 1992 created the threat of deregulation and competition, resulting in efficiency efforts leading to increased equipment loading and less frequent maintenance activities. Details vary for each utility, but the effects of aging on certain classes of aging distribution equipment have begun to manifest themselves in recent years. Mitigating the adverse impact of aging infrastructure and equipment deterioration will require increased levels of investment. The problems associated with aging infrastructure are compounded by increased equipment loading. Not only does increased loading accelerate equipment deterioration. Increased equipment loading also limits the ability of a utility to restore service to customers after a fault occurs. Typically, many interrupted customers can be transferred to another feeder through “back ties.” Presently, many utilities are not able to operate in this manner since performing these transfers, as a result of higher loading, will now result in equipment overloading. Fixing this problem will generally require an increase in the number of feeders and/or back tie locations, with both options requiring increased levels of investment. For many utilities, problems associated with aging substations infrastructure began around 1999. In the last several years many utilities are starting to experience problems with aging cables. Many utilities are also concerned with aging populations of wood poles, but significant problems related to aging wood poles have generally not occurred so far in U.S. utilities. Utilities are recognizing that aging distribution infrastructure will require increased levels of investment. This is one reason that the number of U.S. rate case filings, and the aggregate amounts requested, have significantly increased in recent years. Ratepayer advocates can and should insist that spending related to aging infrastructure be reasonably justified, but limiting a utility’s ability to manage its aging equipment will eventually result in lower levels of reliability and higher cost.
18-31
(PG&E-4)
4 AGING INFRASTRUCTURE AT PG&E This section discusses the state of aging distribution infrastructure at PG&E. It begins by presenting a general overview, and then discusses specific equipment classes in more detail including major substation equipment, underground cables, and wooden poles. In addition, the related topic of back tie capacity is discussed. This section ends with a summary of major findings and opinions.
4.1
Overview of Aging Infrastructure at PG&E
Based on the survivor curves shown in the previous section, distribution equipment at PG&E is older than typical peer utilities. This primarily is a result of two factors. The first is that much of the PG&E service territory, and especially in the San Francisco area, was developed and electrified a long time ago. The second is that much of the PG&E service territory has a mild climate. Population statistics for the twenty largest U.S. cities in 1930 are shown in Table 2.1 (source: U.S. Bureau of the Census). At this point in time, San Francisco ranked eleventh in the nation, with electricity infrastructure construction roughly occurring in proportion to population. Consequently, few cities have as much very old equipment as San Francisco. In addition, the population of San Francisco has not substantially grown since 1930, with the 2000 U.S. census showing a population of 776,733. Based on these demographic factors, San Francisco can be expected to have old distribution equipment, and a high percentage of old distribution equipment. Distribution equipment at PG&E tends to become older than equipment at many other utilities due to a generally mild environment. Distribution equipment at PG&E is not subject to many non-age-related factors of condition deterioration such as periodic hurricanes, severe lightning storms, ice storms, and corrosion due to the use of salt on roadways. When these factors are present, distribution equipment will often fail before becoming very old. For example, major hurricanes periodically destroy many wooden poles in Florida. Consequently, utilities in Florida have a far smaller percentage of wood poles become to become very old when compared to California utilities. The fact that PG&E has an older population of distribution equipment when compared to most peer utilities does not imply that an older population of distribution equipment necessarily requires action at this time. To make a judgment that action is needed now requires a more detailed examination of specific categories of equipment. This is done in the following sections.
18-32
(PG&E-4)
Table 4.1. Largest U.S. Cities in 1930 Rank U.S. City Population in 1930 1 New York, NY 6,930,446 2 Chicago city, IL 3,376,438 3 Philadelphia, PA 1,950,961 4 Detroit, MI 1,568,662 5 Los Angeles, CA 1,238,048 6 Cleveland, OH 900,429 7 St. Louis, MO 821,960 8 Baltimore, MD 804,878 9 Boston, MA 781,188 10 Pittsburgh, PA 669,817 11 San Francisco, CA 634,394 12 Milwaukee, WI 578,249 13 Buffalo, NY 573,076 14 Washington, DC 486,869 15 Minneapolis, MN 464,356 16 New Orleans, LA 458,762 17 Cincinnati, OH 451,160 18 Newark, NJ 442,337 19 Kansas City, MO 399,746 20 Seattle, WA 365,583
4.2
Land Area (sq. mi.) 299.0 201.9 128.0 137.9 440.3 70.8 61.0 78.7 43.9 51.3 42.0 41.1 38.9 62.0 55.4 196.0 71.4 23.6 58.6 68.5
Density (pop. per sq. mi.) 23,179 16,723 15,242 11,375 2,812 12,718 13,475 10,227 17,795 13,057 15,105 14,069 14,732 7,853 8,382 2,341 6,319 18,743 6,822 5,337
Substations
As stated previously, substations generally receive power from one or more transmission lines, step voltage down to primary distribution levels, and supply a circuit breaker at the beginning of each feeder. From a distribution aging infrastructure perspective, the most critical pieces of equipment are substation power transformers and distribution circuit breakers, with most aging infrastructure problems occurring in old substations located in older cities. Age histograms and survivor curves for PG&E substation transformers and substation distribution circuit breakers are shown in Figure 4.1 and Figure 4.2, respectively. These figures are based on data supplied for the KEMA report “2004 Reliability Performance Assessment of San Francisco,” and therefore only include data from the San Francisco operating division. In these figures, the squares correspond to the percentage of equipment at each age, which corresponds to the left vertical axis. This solid, monotonically increasing lines correspond to the percentage of equipment less-than-or-equal-to each age, and corresponds to the right vertical axis. These figures show that over half of the PG&E transformers are over 50 years in age, and about 40% of distribution circuit breakers are over 40 years in age.
18-33
20%
100%
18%
90%
16%
80%
14%
70%
12%
60%
10%
50%
8%
40%
6%
30%
4%
20%
2%
10%
0%
% This age or older .
% Exactly This Age .
(PG&E-4)
0% 0
5
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
Age of Substation Transformers in Years
20%
100%
18%
90%
16%
80%
14%
70%
12%
60%
10%
50%
8%
40%
6%
30%
4%
20%
2%
10%
0%
% This age or older .
% Exactly This Age .
Figure 4.1. The squares connected by the blue line represent the percentage of substation transformers at each specific age. The red line is a survivor curve and shows the percentage of transformers that is older than each age level. These curves show that PG&E has a large number of transformers that are 56 years in age, and that over half of all transformers are older than 50 years.
0% 0
5
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
Age of Substation Breakers in Years
Figure 4.2. The squares connected by the blue line represent the percentage of circuit breakers (distribution class) at each specific age. The red line is a survivor curve and shows the percentage of circuit breakers that is older than each age level. These curves show that PG&E has a large number of transformers that are 56 years in age, and that over half of all transformers are older than 50 years.
18-34
(PG&E-4) To manage its aging substation equipment, PG&E has a sophisticated substation asset management (SAM) program that KEMA considers best practice in the industry. SAM, which has been in place since 1993, allows for centralized decisions with regards to equipment inspection, maintenance, and replacement. Related best-in-class practices include loading guidelines, and the use of mobile spares. Based on the KEMA report “2004 Reliability Performance Assessment of San Francisco,” KEMA has determined that PG&E has a very well-defined equipment replacement program, and this program is successful in identifying and replacing aging equipment, in many instances just prior to failure. For example, 2002 and 2003, two of three transformers that failed in-service were previously identified for replacement. In 2004, seven of nine transformers that failed were targeted for replacement. Few utilities have a program as formal and clearly defined as PG&E’s program.
However, PG&E has not been not replacing equipment at a rate that exceeds its aging rate. This means that the average age of substation transformers and distribution circuit breakers is increasing. For PG&E, equipment chronological age is not a direct factor in equipment replacement, but it is strongly correlated with other condition indicators such as dissolved gas-in-oil analysis. In any case, the PG&E SAM program has, in recent years, been accurately identifying equipment for replacement. However, much of the equipment targeted for replacement has not been replaced due to budget limitations. Of course, every program within a utility must operate within its budget limitations, and SAM has been efficiently allocating its available budget. Over the period of 1996 through 2000 the planned activities accounted for 60% of all activities and the emergency activities accounted for 40%. In the subsequent period (2001 through 2004) this ratio is reversed, with emergency activities accounted for 60% of all activities. It is KEMA’s opinion that this trend is largely due to an inability to replace old equipment that has been identified for replacement. Based on this observation, SAM planned replacement activity should see a large increase when compared to the 2001 through 2004 period.
4.3
Underground Cables
The underground cable situation at PG&E is complex. PG&E has a wide variety of underground cable installations in terms of voltage class, insulation type, conductor size, installation type, and age. For all the complexity, one fact is clear – a lot of underground cable at PG&E is old. Figure 4.3 shows the age histogram and survivor curve for all underground distribution cable. About 20% of all distribution cable is older than 30 years in age, and a noticeable amount is older than 40 years in age. The impact of age on cable condition is becoming evident. Figure 4.4 shows twenty years of recorded outage levels for PILC cable. Since 1983, the number of recorded PILC outages has increased by a factor of seven or more.
18-35
5.0%
100%
4.5%
90%
4.0%
80%
3.5%
70%
3.0%
60%
2.5%
50%
2.0%
40%
1.5%
30%
1.0%
20%
0.5%
10%
0.0%
% This age or older .
% Exactly This Age .
(PG&E-4)
0% 0
5
10
15
20
25
30
35
40
45
50
Age of Cable in Years Figure 4.3. The squares connected by the blue line represent the percentage of underground distribution cable at each specific age. The red line is a survivor curve and shows the percentage of underground distribution cable that is older than each age level. These curves show that about 20% of all distribution cable is older than 30 years in age.
90
# of Lead Outages
80
Lead Outages
70 60 50 40 30 20 10 0 1983
1988
1993
1998
2003
2008
Year
Figure 4.4. This graph shows the number of PILC Cable Outages from 1983 to 2004. Since 1983, the number of recorded PILC outages has increased by a factor of seven or more.
18-36
(PG&E-4)
KEMA has performed an extensive assessment of aging cables at PG&E and has presented its findings in the report “PG&E Aging Cable Modeling and Scenario Analysis”. Based on the results of this study, PG&E can expect the number of cable failures to dramatically increase in the coming years. Presently, PG&E experiences about six hundred cable failures every twelve months. Based on a statistical analysis of PG&E data, this number could increase by a factor of ten over the next thirty years. Currently, an average PG&E customer will experience about 167 interruption minutes per year, with 13 minutes due to cable outages. KEMA predicts that this number will rise from 167 minutes to more than 280 minutes unless some sort of proactive cable replacement is pursued. Statistical analysis, data sufficiency, and materiality have led to the creation of age-based failure rate models for the following classifications of cable:
Statistically Selected Cable Failure Models - Paper Insulated Lead Covered (PILC) - High Molecular Weight Polyethylene (HMWPE) - Cross-Linked Polyethylene rated for 12-kV operation (XLP 12-kV) - Cross-Linked Polyethylene rated for 25-kV operation (XLP 25-kV) - Ethylene Propylene Rubber rated for 12-kV operation (EPR 12-kV) - Ethylene Propylene Rubber rated for 25-kV operation (EPR 25-kV) With the possible exception of PILC, all of these models show that cable failure rates increase exponentially with cable age. For each model, a lower bound, upper bound, and best estimates are provided based on data uncertainty and assumption uncertainty. An example for XLP 12-kV cable is shown in Figure 4.5. KEMA has developed a simulation tool that models cable systems as they age over time and experience failures. Using the above-described cable models, this system is able to predict the number of cable failures in future years, and the resulting impact on reliability indices. Simulation results are shown in Figure 4.6 for the scenario where cables are only replaced as they fail (the model assumes 400 feet of cable replacement per outage). Presently, PG&E experiences about 600 cable failures per year. The graph shows that this number could dramatically increase over the next 15 years, and could peak at more than 7000 failures per year by 2035.
18-37
(PG&E-4)
Failure Rate of XLP 12-kV Cable (/mile/year)
0.7
Historical Failure Rate Best Guess Lower Bound Upper Bound
0.6
Upper Bound
0.5
Best Guess 0.4 0.3
Lower Bound
0.2 0.1 0.0 0
10
20
Age
30
40
50
Figure 4.5. This figure shows the failure rates of XLP 12-kV cables. The three curves represent a lower bound, upper bound, and best guess based on data uncertainty and assumption uncertainty. This type of cable can be expected to have a dramatic increase in the likelihood of failure as it transitions from 30 years in age to 40 years in age.
Figure 4.6. This graph shows the increasing number of cable failures that will occur over time if cables are only replaced after they fail. Presently, PG&E experiences about 600 cable failure per year. Under this scenario, this number could increase to more than 7,000 by the year 2035. Since cable failures tend to impact a large number of customers for a long duration, reliability indices could be severely impacted.
18-38
(PG&E-4) Proactive cable replacement has the potential to mitigate the impact of aging cables. A set of proactive replacement strategies has been examined by KEMA ranging from 100 to 800 circuit miles of proactive replacement per year. These analyses show that about 400 miles of proactive replacement per year is required over the next 35 years in order to prevent a noticeable worsening of customer reliability indices (this number can be lower if more complete cable installation data is available). This amounts to about 14,000 circuit miles of cable replacement over the 35 year period, which will effectively result in a rejuvenated cable system. KEMA in no way recommends this or any other cable replacement strategy. The scenarios were explored so that PG&E can make better informed decisions with respect to aging cables. The KEMA cable study only considered cable on radial parts of the distribution system. By definition, a cable outage on a radial system will interrupt all downstream customers. PG&E also has distribution-class cable that is non-radial. This situation is primarily seen where one substation serves as a supply to another substation through a set of “tie cables.” Tie cables are typically designed so that a substation can be fully supplied with electricity with a single tie cable out of service. Although the failure of a single tie cable will not generally lead to any customer interruptions, the failure of multiple tie cables could lead to a very large number of interruptions. The probability of two tie cables outages occurring at the same time increases with the square of cable failure probability. If tie cable failure rates increase by a factor of five, the probability of customer interruptions will increase at least by a factor of twenty-five. For this reason, deteriorated tie cables at PG&E should be given a high priority for proactive replacement. Presently, underground cable investment decisions are made within the Underground Asset Management program (UAM). Proposed replacement activity for the 2007 filing include tie cable replacement for circuits using PILC and gas-filled cables, vertical runs of PILC cable in indoor substations, cables with severe concentric neutral deterioration, cables with a history of multiple failures, and cables on the “Equipment Requiring Repair” list (ERR). All of these activities are prudent to pursue in the short-run, and address well-known and documented problems. The proposed replacement activity for the 2007 is acceptable in the short term, but proactive cable replacement must eventually increase to prevent a dramatic reduction in customer reliability. However, proactive cable replacement is expensive and is most cost-effective when based on comprehensive cable installation and failure data. PG&E is currently undertaking a Cable Validation Project, which will result in a circuit-by circuit validation of cable section information in the PG&E cable database (C-EDSA). Presently, the contribution of cable failures to customer reliability indices is lower at PG&E than for peer utilities. This is an indication that PG&E has been adequately managing its cables, and that cable failures are not an existing problem. However, the age profile of PG&E cables and the tendency of cable failure rates to increase exponentially with age raises concerns about aging cables in near future.
18-39
(PG&E-4)
4.4
Wood Poles
PG&E has full or joint ownership of about 2.3 million poles, with more than 99 percent of them made of wood. Many of these poles were installed in the 1940s and 1950s, resulting in a large population of old poles. Figure 4.7 shows that about twenty percent of poles are older than fifty years in age, and about three percent of poles are older than eighty years in age. This is concerning since the average service life of a pole is generally about fifty years, and few poles can be expected to have a useful service life greater than eighty years. Unfortunately, it is extremely difficult to predict the expected amount of future useful life for older poles. The best that can generally be done is to periodically inspect poles, and take actions based on these inspections.
5.0%
100%
4.5%
90%
4.0%
80%
3.5%
70%
3.0%
60%
2.5%
50%
2.0%
40%
1.5%
30%
1.0%
20%
0.5%
10%
0.0% 0
10
20
30
40
50
60
70
80
90
% This age or older .
% Exactly This Age .
PG&E has a Pole Asset Management Program that tests and treats its distribution wood poles on a continuous 10-year cycle. In 2004, the program completed its first 10-year cycle and started its second cycle in 2005. All poles receive a visual inspection combined with a “sound and bore” examination. The visual inspection verifies the pole characteristics (pole height, class, wood type, etc.) and identifies any above ground line problems that could render the structure unserviceable. The “sound and bore” test involves hammering completely around the pole, and then bored at several locations on the pole. The bores are probed with an indicator to determine the extent of any voids in that pole from fungal decay or insect infestation. Fumigant (metal-sodium) is then poured into the holes, and the holes are sealed with plugs. In addition, poles are fully or partially excavated, to determine the condition of the pole below groundline. This type of program is typical for large U.S. utilities.
0% 100
Age of Wood Poles in Years
Figure 4.7. The squares connected by the blue line represent the percentage of poles at each specific age. The red solid line is a survivor curve and shows the percentage of poles that is older than each age level. These curves show that about 20% of all wood poles are older than 50 years in age.
18-40
(PG&E-4)
If the pole does not pass inspection, it is considered suspect and requires further analysis. Results of this analysis may result in a recommendation to reinforce the pole through stubbing, or to replace the pole completely. The Pole Stubbing Program is intended to restore the strength of poles decayed at or near the groundline to serviceable condition, extending the poles' serviceable life. Steel trusses (stubs) are the most common form of pole restoration. When a pole is restored through stubbing, the wood is treated with preservatives and a U-shaped steel truss is driven into the ground beside the original pole. These trusses, attached using lag screws, prevent continued scraping of the pole skin by vehicles (tractors, trucks, etc.). The Pole Replacement Program is intended to replace poles for several reasons including: poles that fall below minimum allowable safety factors; poles previously identified for replacement and tracked in the Electric Preventive and Corrective Maintenance (EPCM) database and are flagged as due for replacement in that year; poles along confirmed telecommunications company routes, where the existing or proposed load will cause the pole to have less than the minimum allowable safety factor; poles where the upper pole condition requires action and restoration is not practicable; and poles where the structural integrity may meet minimum standards but has fallen substantially below original design strength. These poles are sorted into a priority list based on shell thickness, circuit criticality, and other factors. In the last several years, PG&E has been replacing about eight thousand poles per year. This level corresponds to a replacement cycle of nearly three hundred years, which is not sustainable in the long term. For the 2007 filing, PG&E is suggesting to replace fifteen thousand poles per year. This corresponds to a replacement cycle of about one hundred and fifty years, which is still likely to be insufficient in the long term. KEMA emphasizes that the impact of aging on pole condition has not become a major problem to date, nor are there signs that problems will emerge in the short term (this is true for both PG&E and for the industry as a whole). However, all old poles will eventually have to be replaced, and the requested level of fifteen thousand poles per year will eventually have to increased at PG&E. Until this occurs, it is important that PG&E continues its test and treat program and begin to use inspection results to identify major trends related to aging.
4.5
Back Tie Capacity
After a fault occurs on a radial distribution system, interrupted customers can be restored through switching procedures. First, the fault is isolated by opening surrounding switches. Next, customers upstream of the isolated area (towards the substation) can be restored from their normal source of power. It is also possible to restore customers downstream of the isolated area by transferring them to other feeders. This is done by closing normally-open tie switches, and is sometimes called “back feeding.”
18-41
(PG&E-4) For back feeding to occur, the load transfer must not result in any equipment overloading. For example, “Feeder A” might have the ability to serve 10 MVA without overloading any equipment. Under normal circumstances, this feeder might serve 8 MVA of load. If a fault occurs on an adjacent feeder, Feeder A can back feed up to 2 MVA without violating loading constraints. If switching limitations allow only a 3-MVA section of customers to be transferred onto Feeder A, the back tie capacity is insufficient for the transfer and the customers must remain interrupted. Aging infrastructure is related to back feeding in several ways. First, aging infrastructure will generally result in more equipment outages, increasing the need for load restoration through back feeding. Second, old equipment is often the limiting factor in back feed capacity (e.g., old wires that tend to be smaller than new conductors). Third, systems with aging infrastructure also tend to be the most heavily loaded. This results in a limited ability to perform load restoration, especially during peak load conditions when equipment failures are most likely. For distribution systems, aging infrastructure problems and back tie capacity often go hand-inhand. Consider Figure 4.8, which shows peak loading levels for feeders originating from some of the older substations in San Francisco. These feeders are ranked according to their 2003 peak load measured as a percentage of both normal summer rating and emergency summer rating. In 2003, PG&E reports that 3.9% of feeders in San Francisco experienced peak loads that exceeded their emergency rating. Figure 4.8 also shows that about twenty percent of feeders from these substations experienced peak loads that exceeded 100% of normal rating. These loading levels are high, and will tend to limit the ability to back feed after a fault occurs.
140% Normal
2003 Peak Load .
120%
Emergency
100% 80% 60% 40% 20% 0% 0
10
20
30
40
50
60
70
80
90
100
Feeder (ranked by emergency loading)
Figure 4.8. 2003 Peak loading for Mission, Larkin, Embarcadero, Potrero, Hunter’s Point and Martin (Feeder 1 is loaded the least as a percentage of its emergency rating and Feeder 100 is loaded the highest as a percentage of its emergency loading).
18-42
(PG&E-4)
KEMA has examined feeder loading data provided by PG&E as it relates to back tie limitations during peak loading conditions. Based on this assessment, KEMA has determined that more than half of all tie switches are limited in their ability to back feed during peak load conditions. These results imply that an increase in back tie capacity will improve system reliability
4.6
Summary
Distribution equipment at PG&E is getting older both in terms of average age and the percentage of very old equipment. The reasons for these trends are the same as for the industry as a whole, but the situation at PG&E is worse than average due to the particular nature of the PG&E service territory (e.g., old development in San Francisco, mild climate, etc.) To date, the effects of aging infrastructure have not had a large impact on customer reliability. However, there are recent signs of equipment deterioration, which means that PG&E will increasingly have to become more proactive in addressing issues related to aging equipment. Specific areas of concern include major substation equipment, underground cables, and wooden poles. In addition, the related topic of back tie capacity will increasingly become important. A summary of important points is now provided:
Key Points on Aging Distribution Infrastructure at PG&E - PG&E does a good job at identifying major pieces of substation equipment that are likely to fail, but has not been able to replace all of the equipment that it recommends for replacement. Consequently, a robust aging infrastructure strategy will likely involve a substantial increase in substation equipment replacement. - PG&E can expect to see a dramatic increase in underground cable failures in the future. To prevent customer reliability from being dramatically reduced, PG&E will eventually have to become much more aggressive in proactive cable replacement. In the short term, there are many identified cable problems that PG&E can specifically address. At the same time, PG&E is gathering data on its cables so that this data can be used to make good replacement decisions in the future. - PG&E has a lot of old wood poles, but no negative trends related to aging have been identified. PG&E has a robust test and treat program, and the number of pole replacements proposed for the 2007 filing is a reasonable in the short term. However, the level of replacement will eventually have to increase in order to prevent an increase in the number of wood pole failures. - Older parts of the PG&E distribution are heavily loaded, resulting in a limited ability to back feed customers after a fault occurs. This results in reduced customer reliability, and can be mitigated by an increase in back tie capacity. With regards to aging infrastructure, PG&E is not perfect. The substation asset management program is best-in-class, but recommendations often are not fully executed. Cable replacement in recent years has been too low. Feeders are very heavily loaded, which will result in accelerated equipment deterioration and a limited ability to restore customers after an outage. These observations show that PG&E has historically made decisions tending towards lower spending, which is not necessarily bad since there must always be a balance between system performance and
18-43
(PG&E-4) spending. However, PG&E’s historical approach is not sustainable since it will result in inadequate system performance and higher total cost of equipment ownership.
PG&E generally spends infrastructure-related investments efficiently, which will be discussed in more detail in the next section. However, the reality is that aging infrastructure will require an increase in spending in the short, medium, and long term. This spending is required both to reduce the life cycle cost of equipment ownership, and to continue providing safe and reliable service to customers.
18-44
(PG&E-4)
5 PG&E APPROACH TO ASSET MANAGEMENT Asset management strives to make spending decisions based on asset-level data so that all corporate performance and risk objectives are met for the lowest possible cost. Throughout the industry, asset management is generally regarded as the most promising approach for making good decisions about aging infrastructure. This section presents a brief overview of asset management, describes general industry trends, and then examines the PG&E approach to asset management as it relates to aging infrastructure.
5.1
Overview of Asset Management
Asset management is a term derived from the financial industry, where its concepts are applied to investment portfolios containing stocks, bonds, cash, options, and other financial instruments. Fundamental to financial asset management is the trade-off between risk and return. Investors identify acceptable risk, and asset management techniques are used to achieve this level of risk for the highest possible return. Also fundamental to asset management is rigorous decision making based on good data and explicit assumptions. Asset management is a way to justify each dollar that is spent. Many techniques of financial asset management are applicable to infrastructure asset management. The objective of asset management is to make all infrastructure-related decisions according to a single set of stakeholder-driven criteria. The payoff is a set of spending decisions capable of delivering the greatest stakeholder value from the investment dollars available. Typically, utilities adopt an asset management approach to either reduce spending, more effectively manage risks, or drive corporate objectives throughout an organization. Stated simply, asset management is a corporate strategy that seeks to balance performance, cost, and risk. Achieving this balance requires the alignment of corporate goals, management decisions, and technical decisions. It also requires the corporate culture, business processes, and information systems capable of making rigorous and consistent spending decisions based on assetlevel data. The result is a multi-year investment plan that maximizes shareholder value while meeting all performance, cost, and risk constraints.
Goals of Asset Management: - Balance cost, performance, and risk - Align corporate objectives with spending decisions - Create a multi-year asset plan based on rigorous and data-driven processes Seen from this perspective, asset management is an appropriate framework for addressing issues related to aging infrastructure. Aging infrastructure often requires hard trade-offs in system performance, system risk, and investment. Aging infrastructure requires that many field-level decisions be made in accordance with the corporate perspective on cost, performance, and risk. Last, aging infrastructure decisions must look well beyond the budget horizon and must be compelling enough to justify the proactive replacement of in-service equipment.
18-45
(PG&E-4)
In its classical form, asset management separates itself from asset ownership and asset operations. The asset owner is responsible for setting financial, technical, and risk criteria. The asset manager is responsible for translating these criteria into an asset plan. The asset service provider is responsible for executing these decisions and providing feedback on actual cost and performance (risk is determined through variation in performance). This decoupled structure allows each asset function to have focus: owners on corporate strategy, managers on planning and budgeting, and service providers on operational excellence. Asset management is also about process. Instead of a hierarchical organization where decisions and budgets follow the chain of command into functional units, asset management is a single process that links asset owners, asset managers, and asset service providers in a manner that allows all spending decisions to be aligned with corporate objectives supported by asset data. In many ways, asset management requires the same skills and functions that have traditionally been found in electric utilities. This includes robust planning, life-cycle cost minimization, effective maintenance management, managerial accounting, cost-to-benefit analysis, and others. The big difference is that effective asset management coordinates and allocates across all spending categories including capital versus expense, capacity versus reliability, repair versus replace, short-term versus long-term, and so forth. Achieving this requires an increased ability to combine business decisions with engineering decisions. If asset management is about aligning spending decisions with corporate objectives, then the role of the asset manager is, in one sense, to bridge the gap between corporate executives and technical engineers. There are other challenges (e.g., culture, data, software), but a utility asset management program must ultimately be able to “connect the dirty boots to the clean suits.”
5.2
Industry Trends
The concepts of financial asset management have gradually been applied to physical asset management over the last twenty years. This started with property management, and was later adopted by the US defense industry, the chemical industry, and most recently public works such as roads and water supplies. Application of asset management for electric utilities began with deregulation in the UK in the early 1990s, and has become a dominant trend at large investorowned utilities in the United States. In KEMA’s experience, electric utilities most commonly pursue asset management to reduce spending, manage the risks across the company, or drive the corporate objectives throughout the organization. Other common considerations are to identify efficiency gains, improve asset replacement techniques, extend asset lives, reduce negative surprises, and provide rigorous justification for expenditures. Most of these goals apply to aging infrastructure, which is why many utilities are basing their aging infrastructure approach on the principles of asset management. Although the role of asset management is different in each utility, many utilities in North America have formal asset management programs. Examples of these include (the list is not comprehensive): American Electric Power, Baltimore Gas & Electric, BC Hydro, Cinergy, Conectiv, Dominion, Duquesne Light, Exelon, First Energy, Georgia Power, GPU, Hydro One, LIPA, Lou-
18-46
(PG&E-4) isville Gas & Electric, Northeast Utilities, Oklahoma Gas & Electric, PacifiCorp, Pacific Gas & Electric, PPL Electric, Progress Energy, Public Service Electric & Gas, Public Service New Mexico, Puget Sound Energy, Southern California Edison, Toronto Hydro, TXU, United Illuminating, WE Energies, and Xcel Energy. Some of these utilities are further along than others, but the clear trend is towards an increasing number of utilities pursuing asset management. Typically the initial focus on asset management is in one of two areas. Many utilities begin their asset management journey by focusing on project prioritization so that all spending can be justified based on common criteria. Other utilities apply asset management to more efficiently spend their inspection and maintenance dollars. To apply asset management to aging infrastructure, utilities must have capability in both project prioritization and in maintenance management. Few utilities are at a point in the organizational learning process where they can treat aging infrastructure in a comprehensive manner. Many utilities have developed specialized programs designed to manage a specific class of asset (e.g., cables), and many of these programs are sophisticated and effective. However, there are no electric utilities that KEMA is aware of that are able to make all aging infrastructure-related decisions through a centralized, coordinated, and data driven process. Most major stakeholders agree that asset management has potential benefits for many utilities. The IEEE and CIGRE have recently formed a working groups on asset management. EPRI offers an “Asset Management Toolkit.” Many software vendors are enhancing products to support asset management solutions. This includes financial packages (e.g., SAP), geographic information systems, computerized maintenance management systems, and others. Nearly every consulting company that has electric utilities as clients (including KEMA) has asset management offerings. There are several industry seminars each year on T&D asset management. Major industry conferences commonly have at least one panel session devoted to asset management. The point is that asset management is pervasive, and is accepted as a preferred business approach by many utilities, consultants, and industry associations.
5.3
Asset Management at PG&E
PG&E has an Asset Management type of organization with Engineering & Planning functions separate from Operations, Maintenance & Construction. The former is responsible for the planning functions to arrive at an effective project plan. The latter is responsible as a service provider for the efficient execution of those projects. This type of organization is conducive to good investment prioritization since it separates the role of decision making from the role of project execution. Annual budgeting is based on rigorous plan review cycle, including cross-functional project optimization and resource planning. This integrated planning and budgeting process results in the following: a 5-year plan (revised annually); a 6-quarter rolling forecast of activities and expenditures; and a detailed 1-year budget. There is monthly tracking of costs and projects, and a mechanism for early release of “must-do” projects.
18-47
(PG&E-4) PG&E utilizes a program management philosophy to prioritize and budget for major spending decisions. A program is a distinct grouping of related spending activities, represented by a one or more Major Work Categories. Each program has specific targets in order to ensure corporate objectives like safe, reliable and responsive service, and earning the authorized rate of return. Each program is charged to come up with a list of non-discretionary and discretionary projects over a 5-year horizon based on detailed selection criteria. Non-discretionary spending is typically related to safety and compliance (e.g., emergency response, capacity, etc.), customer connection and mandatory projects. Discretionary spending is related to maintaining system performance including distribution reliability and some asset replacement work. After the program reviews, each line of business (Utility Operations, Generation, and the Corporate Services Departments) prioritizes expenditures across its programs. Then, all program budget requests and prioritization of expenditures are reviewed together by a Budget Working Group, led by the budget department director and consisting of representatives from each line of business to ensure consistency across programs company-wide. Finally, the Senior Officer team reviews the proposed budgets and any anticipated shortfall in revenues, and prioritizes expenditures across the Utility based on considerations including safety, reliability, customer service and financial goals to finalize a budget recommendation and to formulate contingency plans. In terms of aging infrastructure, key programs include substation asset management (SAM), underground asset management (UAM), and pole asset management (PAM). Each of these programs with respect to asset management is summarized as follows:
Substation Asset Management. SAM does an effective job at assessing equipment condition, specifying cost-effective maintenance policies, and identifying equipment that should be replaced. In recent years, budget limitations have preventing SAM from pursuing all of its replacement recommendations. Underground Asset Management. UAM is less mature than SAM at PG&E. This is appropriate since the impact of aging cables has occurred later than the impact of aging substation equipment. UAM is, in the short term, planning to address specific cable issues such as tie cable replacement, vertical PILC cable replacement in substations, replacement of cable with deteriorated concentric neutral conductors, and replacement of cable that has failed multiple times in the past. These are prudent expenditures, but UAM must gradually expand its ability to address more general issue of aging cables and the resulting increase in cable outages. UAM recognizes this need and has undertaken an aggressive data collection and validation program. Pole Asset Management. The PAM group is similar in function to other peer utilities, and is able to effectively allocate spending with regards to aging wood poles. To date, the PAM group has not looked at macro trends in wood pole performance. This is acceptable since wood pole failures have not been a large concern. However, PAM must eventually become more sophisticated in analyzing test and treat data so that it can track the inevitable condition deterioration of aging wood poles. Based on its asset management organization and processes, PG&E is capable of making efficient spending decisions related to aging substation equipment, aging underground cable, and aging
18-48
(PG&E-4) wood poles. In the future, UAM and PAM must become more sophisticated for efficient spending to remain in effect. PG&E is also prepared to make efficient spending decisions with regards to back tie capacity through its Reliability Program. Within this program, reliability-related investments that cannot be classified as mandatory must be analyzed using a benefit-to-cost ratio (BCR). The BCR is determined by dividing the net present value (NPV) of a project’s estimated customer value by the NPV of the costs of the project. The estimated customer value is intended to represent a customer value of service (VOS). Values used to compute VOS were derived from PG&E’s 2000 performance-based rate-making filing. Not only does this approach allow for efficient spending with regards to back tie capacity, but allows these projects to be compared against other capital projects with associated reliability improvements. PG&E also has a process in place to coordinate spending across programs. This process is nonrigorous in the sense that it is not able to objectively rank all projects in all programs. There are several compelling reasons for this that are not unique to PG&E. Some of the major difficulties in applying traditional project ranking approaches to aging infrastructure include: -
-
-
-
-
Multiple performance measures are difficult to compare. For example, one project may improve safety, while another project may improve overall system reliability, and a third project may address the specific concerns of a major customer. It is difficult to prioritize these projects in a rigorous manner. Ideally, the NPV of a project should consider the entire life cycle cost of the project including operations, inspection, maintenance, and retirement. In terms of aging infrastructure, there is large uncertainty in terms of remaining life and escalating O&M costs. Within a program, there are often pragmatic reasons that require overall capital and expense budgets to be met, which affect the proposed project mix submitted for approval. This list could potentially be different without CAPEX versus OPEX considerations. This makes it difficult to recognize when, for example, one program would prefer to spend more on CAPEX and another would prefer to spend more on OPEX, leading to sub-optimization. Many of the most important aging infrastructure concerns deal with low probability events that have high impact. Many of these “headline events” do not have a large impact on traditional performance measures, but can have a large cost or regulatory or customer satisfaction impact that is easily quantified. These risk considerations are real, can easily dominate spending priority, but are generally treated in a qualitative manner. Urban areas will typically have better reliability performance when compared to non-urban areas. This is due to a variety of factors including load density and a high percentage of underground equipment. Consider an urban area with a SAIDI of 60 minutes and a rural area with a SAIDI of 300 minutes. Now consider a program that targets the reliability of the worst 10% of customers in the city and another program that targets the reliability of the worst 10% of customers in rural area. If each program improves SAIDI by 10% for its area, the benefitto-cost ratio of the city program will look worse by a factor of five, simply because initial SAIDI is five times lower. This makes it more difficult to justify spending on aging infrastructure in urban areas, which is precisely where most aging infrastructure exists.
18-49
(PG&E-4)
The above difficulties show that traditional approaches to project prioritization are limited in their ability to address certain issues related to aging infrastructure, especially with regards to cross-program budget allocation. Given these difficulties, PG&E’s approach to asset management and aging infrastructure is appropriate, both in terms of timing and in terms of the ability to make prudent decisions. Specifically, SAM was created soon after substation aging infrastructure started to present difficulties. UAG was created soon after PG&E cable aging infrastructure started resulting in a substantial number of increased outages. PAM has been in place for ten years and is comparable to best industry practice. PG&E is also pursuing asset management at the power delivery level, which will, over time, help to address aging infrastructure difficulties. The last issue related to asset management is risk management, and the related topic of “gold plating.” Risk is related to the random nature of future performance. For utilities, many decisions are based on expected future performance. For example, a project may be justified based on its expected reliability benefits or its expected reduction in operating costs. Utilities are also interested in minimizing the probability of poor outcomes. Consider a utility with a SAIDI of two hours. Each year, this number could be higher or lower simply due to good luck or bad luck. A utility may choose to improve the predictability of SAIDI rather than improve average SAIDI. Managing the risk of bad financial outcomes is a well-defined field called “value at risk.” Investment banks may be required to measure, for example, the worst one-in-ten outcome for an investment portfolio. Many utilities have analogous criteria in generation planning (plan for the worst one day in ten years) and distribution capacity planning (plan for the worst one-in-ten year weather). However, many aspects of risk are difficult to treat in a rigorous manner due to a lack of historical data and a lack of predictive modeling. It must be emphasized that spending based on “risk mitigation” is common practice and cannot be equated to “gold plating.” For example, engineering calculations can determine the most economical pole size based on pole strength and the weight of the equipment that will be placed on the pole. However, not all poles have identical strength. Poles gradually lose strength over time. Additional equipment may (or may not) have to be at a particular location in the future. Based on these uncertainties, a stronger pole may be specified to ensure safety and to deliver better economy across a range of future possible scenarios. Justifying projects based on risk factors can be abused, but KEMA has not seen this behavior at PG&E. To the contrary, PG&E attempts to base decisions based on rigorous analysis, which tends to result in lower spending on risk mitigation. This opinion is similar to the results found in the 2002 report “Technical Audit of PG&E’s 1999 Capital Expenditures,” where Stone & Webster scored projects in each major work category (MWC) based on “necessity, reasonableness, effectiveness, and the prudence of the investments.” The report concluded “lower scores correspond to MWC’s associated with Work Requested by Others, and the higher scores are associated with capacity additions or capitalized maintenance activities (Maintenance & Reliability).”
18-50
(PG&E-4) As the need for investment in aging infrastructure increases, PG&E will need to become more focused on risk-based spending. All spending decisions should be as data-driven and rigorous as is practical, but risk-based spending will always be more subjective than performance-based spending since there are more areas of uncertainty.
5.4
Summary
Overall, reliability investment prioritization at PG&E is equal-to or better-than typical large investor-owned utilities in the U.S. This includes organization, project ranking, budgeting, and cross-program optimization. There is opportunity for improvement, but the capabilities of PG&E are sufficient to make good decisions with regards to present aging infrastructure issues. In addition, PG&E is taking steps to improve this area, which should allow for a continued ability to make good spending decisions as aging infrastructure issues become more complex over time.
18-51
(PG&E-4)
6 CONCLUSIONS Aging infrastructure is becoming an increasing concern for the PG&E distribution system. As such, PG&E should not delay implementation of its proposed practices in the areas of substation equipment, underground cable, and wood poles. KEMA feels that proposed spending in these areas is conservative, and increased levels of spending are justified. Aging infrastructure issues at PG&E will get worse before they get better, and PG&E will have to increasingly spend money in these areas as the aging process continues to worsen the condition of equipment. In each area, KEMA has examined the quality of PG&E decision-making in terms of safety, reliability, and future costs. Safety levels must be adequate; reliability must be balanced with cost; and life-cycle costs should be minimized. Overall, reliability investment prioritization at PG&E is equal-to or better-than typical large investor-owned utilities in the U.S. This includes organization, project ranking, budgeting, and cross-program optimization. There is opportunity for improvement, but the capabilities of PG&E are sufficient to make good decisions with regards to present aging infrastructure issues. In addition, PG&E is taking steps to improve this area, which should allow for a continued ability to make good spending decisions as aging infrastructure issues become more complex over time. Comments on specific issues are now provided.
Substation Equipment. PG&E is effective at identifying major pieces of substation equipment that are likely to fail. This is based on a mature Substation Asset Management program (SAM) that has been in place since 1993. However, this program has not been able to replace all of the equipment that it recommends for replacement due to budget limitations. Consequently, a robust aging infrastructure strategy must involve increases in substation equipment replacement spending. Underground Cables. PG&E can expect to see a dramatic increase in underground cable failures in the future. To prevent customer reliability from being significantly reduced, PG&E must eventually become more aggressive in proactive cable replacement. In the short term, PG&E proposes to address specific problems, and to build a comprehensive database so that broader decisions can be made in the future through its Underground Asset Management (UAM) program. This approach is conservative and justified. Wood Poles. PG&E has a substantial number of very old wood poles; to date, no negative trends related to aging have been identified. However, the number of pole replacements in recent years is unsustainable, even by conservative estimates. The PG&E proposed increase in pole replacements is conservative and justified. Back Tie Capacity. Older parts of the PG&E distribution system are heavily loaded, resulting in a limited ability to reconfigure the system and restore customers after a fault occurs (referred to as Back Tie Capacity). PG&E has processes in place to justify reliability improvement spending based on cost-to-benefit analysis. PG&E needs to increase its back tie capacity on many feeders, and back-tie-capacity projects selected through these processes are justified. Many other utilities are recognizing that aging distribution infrastructure will require increased levels of investment. This has resulted in a large increase in the number of recent U.S. rate case
18-52
(PG&E-4) filings, and in the aggregate amounts requested in these filings. Often these increases are contested by ratepayer advocacy groups, which strive for adequate utility performance for the lowest possible rates. KEMA understands the need for ratepayer advocacy, but has the opinion that increases in distribution aging infrastructure spending at PG&E are required to both ensure adequate service and to minimize the life cycle cost of equipment ownership With regards to aging infrastructure, PG&E is not perfect. The substation asset management program is best-in-class, but recommendations often are not fully executed. Cable replacement in recent years has been too low. Feeders are very heavily loaded, which will result in accelerated equipment deterioration and a limited ability to restore customers after an outage. These observations show that PG&E has historically made decisions tending towards lower spending, which is not necessarily bad since there must always be a balance between system performance and spending. However, PG&E’s historical approach is not sustainable since it will result in inadequate system performance and higher total cost of equipment ownership. The PG&E planning and budgeting processes are able to identify good spending decisions as they relate to aging distribution equipment. These processes have successfully identified areas requiring increased levels of spending, have successfully identified specific projects that should be implemented in the short term, and have successfully identified data that must be collected so that efficient decisions can continue to be made. This said, aging infrastructure issues at PG&E will get worse before they get better, and will require an increase in spending in the short, medium, and long term. This spending is required to reduce the life cycle cost of equipment ownership, to ensure safety, and to continue providing adequate levels of service to customers.
18-53
(PG&E-4)
APPENDIX A. Qualifications of Richard E. Brown Contact Information 3801 Lake Boone Trail, Suite 200 Raleigh, NC 27607 Voice: (919) 593-2860 Fax: (215) 822-4267 E-mail:
[email protected] Summary of Qualifications Richard Brown is a senior principal consultant with KEMA. He is a recognized international expert in power system reliability and asset management, has published more than 70 technical papers and articles in these areas, and is author of the book Electric Power Distribution Reliability. Dr. Brown is a senior member of IEEE and a registered professional engineer. Education Degree M.B.A. Ph.D. M.S.E.E. B.S.E.E.
Institution University of North Carolina University of Washington University of Washington University of Washington
Professional Experience Title Senior Principal Consultant Director of Technology Principal Engineer Senior Engineer Research/Teaching Assistant Electrical Engineer III Electrical Engineer II
Location Chapel Hill, NC Seattle, WA Seattle, WA Seattle, WA
Institution KEMA Inc. ABB Consulting ABB Power Distribution Solutions ABB Corporate Research University of Washington Jacobs Engineering Jacobs Engineering
Year 2003 1996 1993 1991
Dates 5/2003 5/2001 2/1999 7/1996 1/1994 1/1993 4/1991 -
Professional Registration and Professional Societies • Senior Member, IEEE Power Engineering Society - Member, Power System Planning and Implementation Committee (1997-present) - Chair, Distribution Working Group (2003-present) - Chair, Power Delivery Reliability Working Group (1997-1999) - Member, Working Group on System Design (1997-present) - President, University of Washington Student Chapter (1994-1995) - Vice President, University of Washington Student Chapter (1993-1994) - Technical Paper Reviewer for IEEE Transactions (1996-present) • Registered Professional Engineer in the State of North Carolina (Certificate No. 23088) Honors and Awards • IEEE PES Walter Fee Outstanding Young Engineer Award (2003) • Listed in Marquis Who’s Who in America • Listed in Marquis Who’s Who in Science and Engineering • Listed in Madison’s Who’s Who • ABB Award of Excellence: President’s Award (1999), Product Development (1998) • Member: Eta Kappa Nu (Electrical Engineering Honor Society) • Member: Beta Gamma Sigma (Business Honor Society)
18-54
present 4/2003 4/2001 1/1999 6/1996 12/1993 12/1992
(PG&E-4)
Publications and Appearances • Books and Theses: • Refereed Journal Papers: • Refereed Conference Papers: • Technical Articles: • Invited Presentations:
5 16 41 14 36
(Details on Page 57) (Details on Page 57) (Details on Page 58) (Details on Page 60) (Details on Page 60)
Professional Experience 5/03 – present
KEMA T&D Consulting, Raleigh, NC Senior Principal Consultant—As a charter member of the T&D Consulting (TDC) division in the US, my role is to provide management and technical consulting services in the areas of reliability and asset management. I also serve on the US TDC leadership team, manage the asset management group, and am responsible for strategic planning and business development. Since inception, I have played a key role in growing the group to a $4 million per year business with cumulative positive cash flow within 18 months. Key client engagements during this time include Constellation Group, Xcel Energy, Pacific Gas & Electric, Southern California Edison, Cinergy, and Midwest Energy, and many others.
7/96 – 4/03
ABB
5/01 – 4/03
Director of Technology, ABB Consulting, Raleigh, NC ABB Consulting provides technical advice and training for both internal and external customers. As Director of Technology, I had the responsibility for research and development of algorithms and tools to support existing capabilities and create new opportunities. In addition, I had the responsibility for developing an eConsulting platform capable of providing comprehensive consulting services through an application service provider paradigm (e.g., system modeling, data warehousing, knowledge bases, on-line training). Although this position is within the US Consulting business of ABB, its scope was to provide global support for the entire $500 million Utility Partners business area. - Served as a consultant for the following electric utilities: Florida Progress, PacifiCorp, PECO, Xcel, Dominion, Endessa (Spain), Niagara Mohawk, Commonwealth Edison, Consolidated Edison, National Grid, Scottish Power, Tampa Electric Company. - Line manager for a small team of highly skilled R&D engineers.
2/99 – 4/01
Principal Engineer, ABB Power Distribution Solutions, Raleigh, NC Power Distribution Solutions was created by ABB to provide customers with complete solutions based on functional requirements including design, build, own, operate, maintain, guarantee, and finance. I was selected to be a charter member of this group responsible for design expertise, optimization expertise, and software expertise. Major accomplishments include: - Line manager for a small team of highly skilled R&D engineers. - Development of Performance AdvantageTM, ABBs internal distribution system analysis and optimization tool. This tool is capable of optimizing all aspects of distribution systems including electrical performance, reliability, economics, and risk. I was awarded the ABB Award of Excellence (Product Development) for this effort. - Development of Strategic AdvantageTM, ABBs internal spatial load forecasting and substation planning tool. - Played a key role in securing a $127 million solution sale to Commonwealth Edison after their reliability problems in the summer of 1999. The capture team of seven people was awarded the ABB Award of Excellence (President’s Award) for this effort. - Served as a consultant for the following electric utilities: Commonwealth Edison, Carolina Power and Light, TXU, NStar, Scottish Power, PacifiCorp, Florida Power & Light.
18-55
(PG&E-4) - Served as a consultant for the following C&I customers: General Motors, Ford, Monsanto, Armco Steel, Mobil Oil, Chevron. - Instructor for numerous external and internal workshops in the areas of engineering, planning, reliability, and design optimization. 7/96 - 1/99
Senior Engineer, ABB Corporate Research, Raleigh, NC ABBs R&D facility in the United States (formerly called The Electric Systems Technology Institute) is located on the Centennial Campus of North Carolina State University. My job responsibilities included research, product development, consulting, project management, business development, and teaching workshops. Major accomplishments include: - Created distribution system reliability assessment software sold commercially as ReliNETTM. - Created substation reliability assessment software (SUBREL) distributed to ABB substation groups globally. - Created budget constrained planning software that used marginal cost/benefit methods to optimally allocate utility capital and O&M budgets. - Served as project manager for a $610,000 corporate research project that developed the tools and expertise necessary for ABB to transition from an equipment provider to a solution provider. - Provided consulting services for the following electric utilities: Duke Energy, Midwest Energy, Ameren, Meralco (Philippines), AEP, GPU, Florida Power and Light, Georgia Power, Baltimore Gas and Electric, PECO.
1/94 - 6/96
UNIVERSITY OF WASHINGTON, Seattle, WA Research/Teaching Assistant—My research done at the University of Washington was in the area of distribution system reliability assessment and design optimization. Research was funded by Snohomish County PUD #1 on 2 successive contracts totaling $170,000. This project resulted in a distribution system reliability assessment software package (DS-RADS) which was later sold to Power Technologies, Inc. and is now a commercially available product. In addition to research, I served as a teaching assistant for various power systems and controls courses at the undergraduate and graduate level.
4/91 – 12/93
JACOBS ENGINEERING, Kirkland, WA Engineer II and Engineer III—Jacobs Engineering (formerly Sverdrup Corp.) is a multidisciplinary engineering firm with electrical, mechanical, civil, and structural design capabilities. Responsibilities included engineering design of medium voltage and low voltage electrical systems for industrial facilities, institutional facilities, and public works. Typical work included design, value engineering, specification writing, construction document generation, and construction support. Major projects included: - University of Florida (Orlando) Biotechnology Research and Development Facility: Lead electrical engineer including underground service, main switchgear, MCCs, emergency generation, life safety system, exterior lighting and interior lighting. - Boeing Headquarters: Electrical systems design for an office park main substation, central plant, communications building, and underground site distribution system. Duties included design of a 115-kV, 25-MVA substation, protective relaying, 15-kV, 5-kV, and 600-V distribution, and fire detection/alarm systems. Performed an energy conservation study funded by Puget Sound Energy. - Boeing Research Aerodynamic Icing Tunnel: Electrical systems design for a new equipment building, substation expansion, and wind tunnel structure. Duties included 600-V distribution system design and cost estimation for demolition and construction. - Boeing Research Hot Gas Test Facility: Electrical systems design for a three-cell hot gas test facility. Duties included grounding system design, 600-V distribution system design, and heat trace system design.
18-56
(PG&E-4) - Arizona DOT SR-360 Traffic Interchange: Electrical systems design for an outer highway loop including two tunnels. Duties included design of staged interior HID tunnel lighting, 600-V power distribution, signal reference grid, and control room.
Books and Theses 1. 2. 3. 4. 5.
R. E. Brown, Electric Power Distribution Reliability, Marcel Dekker, 2002. R. E. Brown, author of chapter “Distribution System Reliability: Analytical and Empirical Techniques” in IEEE Tutorial on Electric Delivery System Reliability Evaluation, J. Mitra (Editor), IEEE, 2005, pp. 39-51. R. E. Brown, author of chapter “Power System Reliability” in Electric Power Engineering Handbook, L. L. Grigsby (EIC), CRC Press LLC, 2001, pp. 13-51 through 13-65. R. E. Brown, Reliability Assessment and Design Optimization for Electric Power Distribution Systems, Ph.D. Dissertation, University of Washington, Seattle, WA, 1996. R. E. Brown, An Intelligent Overload Relay for Extruded Dielectric Transmission Cable, Masters Thesis, University of Washington, Seattle, WA, 1993.
Refereed Journal Papers 1. 2. 3. 4. 5.
6. 7. 8. 9.
10. 11. 12. 13.
14.
15.
R. E. Brown, M. V. Engel, J. H. Spare, “Making Sense of Worst Performing Feeders”, (accepted for publication in IEEE Transactions on Power Systems). R. E. Brown, G. Frimpong, H. L. Willis, “Failure Rate Modeling Using Equipment Inspection Data”, IEEE Transactions on Power Systems, Vol. 19, No. 2, May 2004, pp. 782-787. S. S. Venkata, A. Pahwa, R. E. Brown, and R. D. Christie, “What Future Distribution Engineers Need to Learn,” IEEE Transactions on Power Systems, Vol. 19, No. 1, Feb. 2004, pp. 17-23. F. Li and R. E. Brown, “A Cost-Effective Approach of Prioritizing Distribution Maintenance Based on System Reliability,” IEEE Transactions on Power Delivery, Vol. 19, No. 1 , Jan. 2004, pp. 439-441. F. Li, R. E. Brown, and L. A. A. Freeman, “A Linear Contribution Factor Model of Distribution Reliability Indices and its Applications in Monte Carlo Simulation and Sensitivity Analysis,” IEEE Transactions on Power Systems, Vol. 18, No. 3, Aug. 2003, pp. 1213-1215. R. E. Brown and A. P. Hanson, “Impact of Two Stage Service Restoration on Distribution Reliability,” IEEE Transactions on Power Systems, Vol. 16, No. 4, Nov. 2001, pp. 624-629. R. E. Brown and J. J. Burke, “Managing the Risk of Performance Based Rates,” IEEE Transactions on Power Systems, Vol. 15, No. 2, May 2000, pp. 893-898. R. E. Brown and M. M. Marshall, “Budget Constrained Planning to Optimize Power System Reliability,” IEEE Transactions on Power Systems, Vol. 15, No. 2, May 2000, pp. 887-892. R. E. Brown, “The Impact of Heuristic Initialization on Distribution System Reliability Optimization,” International Journal of Engineering Intelligent Systems for Electrical Engineering and Communications, Vol. 8, No. 1, March 2000, pp. 45-52. R. E. Brown, J. R. Ochoa, “Impact of Sub-Cycle Transfer Switches on Distribution System Reliability,” IEEE Transactions on Power Systems, Vol. 15, No. 1, Feb. 2000, pp. 442-447. R. E. Brown, T. M. Taylor, “Modeling the Impact of Substations on Distribution Reliability,” IEEE Transactions on Power Systems, Vol. 14, No. 1, Feb. 1999, pp. 349-354. R. E. Brown, J. R. Ochoa, “Distribution System Reliability: Default Data and Model Validation,” IEEE Transactions on Power Systems, Vol. 13, No. 2, May 1998, pp. 704-709. R. E. Brown, S. Gupta, R. D. Christie, S. S. Venkata, and R. D. Fletcher, “Distribution System Reliability: Momentary Interruptions and Storms,” IEEE Transactions on Power Delivery, Vol. 12, No. 4, October 1997, pp. 1569-1575. R. E. Brown, S. Gupta, R. D. Christie, S. S. Venkata, and R. D. Fletcher, “Automated Primary Distribution System Design: Reliability and Cost Optimization,” IEEE Transactions on Power Delivery, Vol. 12, No. 2, April 1997, pp. 1017-1022. R. E. Brown, S. Gupta, R. D. Christie, S. S. Venkata, and R. D. Fletcher, “Distribution System Reliability Analysis Using Hierarchical Markov Modeling,” IEEE Transactions on Power Delivery, Vol. 11, No. 4, Oct. 1996, pp. 1929-1934.
18-57
(PG&E-4) 16. V. N. Chuvychin, N. S. Gurov, S. S. Venkata, and R. E. Brown, “An Adaptive Approach to Load Shedding and Spinning Reserve Control During Underfrequency Conditions,” IEEE Transactions on Power Systems, Vol. 11, No. 4, Nov. 1996, pp. 1805-1810.
Refereed Conference Papers 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
22. 23.
R. E. Brown “Project Selection with Multiple Performance Objectives,” 2005 IEEE/PES Transmission and Distribution Conference and Exposition, New Orleans, LA, Sept. 2005. R. E. Brown, J. Spare, “The Effects of System Design on Reliability and Risk,” 2005 IEEE/PES Transmission and Distribution Conference and Exposition, New Orleans, LA, Sept. 2005. R. E. Brown, J. Spare “A Survey of U.S. Reliability Reporting Processes,” 2005 IEEE/PES Transmission and Distribution Conference and Exposition, New Orleans, LA, Sept. 2005. R. E. Brown and J. Spare, “Asset Management and Financial Risk,” DistribuTECH Conference and Exhibition, San Diego, CA, Jan. 2005. R. E. Brown and J. H. Spare, “Asset Management, Risk, and Distribution System Planning,” IEEE Power Systems Conference and Exhibition, New York, NY, Oct. 2004. R. E. Brown, “Identifying Worst Performing Feeders,” Probabilistic Methods Applied to Power Systems, PMAPS 2004, Ames, IA, September 2004. H. L. Willis, M. V. Engel and R. E. Brown, “Equipment Demographics – Failure Analysis of Aging T&D Infrastructures,” 2004 Canada Power Conference, Toronto, Canada, September 2004. R. E. Brown, “Failure Rate Modeling Using Equipment Inspection Data”, IEEE PES 2004 General Meeting, Denver, CO, June 2004. R. E. Brown, “Coming to Grips with Distribution Asset Management,” 2003 Real World Conference: It’s All About Cost and Reliability, Transmission and Distribution World, Ft. Lauderdale, FL, Oct. 2003. R. E. Brown, “Reliability Standards and Customer Satisfaction,” 2003 IEEE/PES Transmission and Distribution Conference and Exposition, Dallas, TX, Sept. 2003. A. Pahwa, S. Gupta, Y. Zhou, R. E. Brown, and S. Das, “Data Selection To Train A Fuzzy Model For Overhead Distribution Feeders Failure Rates," International Conference on Intelligent Systems Applications to Power Systems, Lemnos, Greece, Sept. 2003. R. E. Brown, “Network Reconfiguration for Improving Reliability in Distribution Systems,” IEEE PES 2003 General Meeting, Toronto, Canada, July 2003. R. E. Brown, , J. Pan, Y. Liao, and X. Feng, “An Application of Genetic Algorithms to Integrated System Expansion Optimization,” IEEE PES 2003 General Meeting, Toronto, Canada, July 2003. R. E. Brown and L. A. A. Freeman, “A Cost/Benefit Comparison of Reliability Improvement Strategies,” DistribuTECH Conference and Exhibition, Las Vegas, NV, Feb. 2003. S. Gupta, A. Pahwa, R. E. Brown and S. Das, “A Fuzzy Model for Overhead Distribution Feeders Failure Rates,” NAPS 2002: 34th Annual North American Power Symposium, Tempe, AZ, Oct. 2002. R. E. Brown, “Web-Based Distribution System Planning,” IEEE PES Summer Power Meeting, Chicago, IL, July 2002. R. E. Brown, “System Reliability and Power Quality: Performance-Based Rates and Guarantees,” IEEE PES Summer Power Meeting, Chicago, IL, July 2002. R. E. Brown, “Modeling the Reliability Impact of Distributed Generation,” IEEE PES Summer Power Meeting, Chicago, IL, July 2002. S. Gupta, A. Pahwa, R. E. Brown, “Data Needs for Reliability Assessment of Distribution Systems,” IEEE PES Summer Power Meeting, Chicago, IL, July 2002. R. E. Brown, “Meeting Reliability Targets for Least Cost,” DistribuTECH Conference and Exhibition, Miami, FL, Feb. 2002. S. Gupta, A. Pahwa and R. E. Brown, “Predicting the Failure Rates of Overhead Distribution Lines Using an Adaptive-Fuzzy Technique,” NAPS 2001: 33rd Annual North American Power Symposium, College Station, TX, Oct. 2001. P. R. Jones and R. E. Brown, “Advanced Modeling Techniques to Identify and Minimize the Risk of Aging Assets on Network Performance,” Utilities Asset Management 2001, London, UK, July 2001. R. E. Brown, “Distribution Reliability Modeling at Commonwealth Edison,” 2001 IEEE/PES Transmission and Distribution Conference and Exposition, Atlanta, GA, Oct. 2001.
18-58
(PG&E-4) 24. R. E. Brown, “Distribution Reliability Assessment and Reconfiguration Optimization,” 2001 IEEE/PES Transmission and Distribution Conference and Exposition, Atlanta, GA, Oct. 2001. 25. R. E. Brown, J. Pan, X. Feng and K. Koutlev, “Siting Distributed Generation to Defer T&D Expansion,” 2001 IEEE/PES Transmission and Distribution Conference and Exposition, Atlanta, GA, Oct. 2001. 26. D. Ross, L. Freeman and R. E. Brown, “Overcoming Data Problems in Predictive Distribution Reliability Modeling,” 2001 IEEE/PES Transmission and Distribution Conference and Exposition, Atlanta, GA, Oct. 2001. 27. R. E. Brown and L. A. A. Freeman, “Analyzing the Reliability Impact of Distributed Generation,” IEEE PES Summer Power Meeting, Vancouver, BC, Canada, July 2001. 28. R. E. Brown and M. Marshall, “Microeconomic Examination of Distribution Reliability Targets,” IEEE PES Winter Power Meeting, Columbus, OH, Jan. 2001, Vol. 1, pp. 58-65. 29. P. R. Jones and R. E. Brown, “Investment Planning of Networks Using Advanced Modeling Techniques,” Utilities Asset Management 2001, London, UK, Jan. 2001. 30. R. E. Brown, “Probabilistic Reliability and Risk Assessment of Electric Power Distribution Systems,” DistribuTECH Conference and Exhibition, San Diego, CA, Feb. 2001. 31. C. LaPlace, D. Hart, R. E. Brown, W. Mangum, M. Tellarini, J. E. Saleeby, “Intelligent Feeder Monitoring to Minimize Outages,” Power Quality 2000 Conference, Boston, MA, Oct. 2000. 32. R. E. Brown, H. Nguyen, J. J. Burke, “A Systematic and Cost Effecting Method to Improve Distribution Reliability,” IEEE PES Summer Meeting, Edmonton, AB, July 1999. Vol. 2, pp. 1037-1042. 33. R. E. Brown, T. M. Taylor, “Modeling the Impact of Substations on Distribution Reliability,” IEEE PES Winter Meeting, New York, NY, Feb 1999, pp. 349-354. 34. R. E. Brown, A.P. Hanson, M.M Marshall, H.L. Willis, B. Newton, “Reliability and Capacity: A Spatial Load Forecasting Method for a Performance Based Regulatory Environment,” 1999 Power Industry Computer Applications Conference, Dayton, OH, February 1999, pp. 139-144. 35. R. E. Brown, A. P. Hanson, D. Hagan, “Long Range Spatial Load Forecasting Using Non-Uniform Areas,” 1998 IEEE/PES Transmission and Distribution Conference, New Orleans, LA, April 1999, Vol. 1, pp. 369-373. 36. R. E. Brown, W. S. Zimmermann, P. P. Bambao Jr., and L. P. Simpao, “Basic Planning for a New Fast Growing Area in Manila with a Total Electrical Load of 650 MVA,” 12th Annual Conference of the Electric Power Supply Industry, Pattaya, Tailand, November 1998. 37. X. Y. Chao, R. E. Brown, D. Slump, and C. Strong, “Reliability Benefits of Distributed Resources,” Power Delivery International ‘97 Conference, Dallas, TX, December 1997. 38. R. E. Brown, “Competitive Distribution Systems: A Reliability Perspective,” American Power Conference, Vol. 59-II, Chicago, IL, April 1997, pp. 1115-1120. 39. R. E. Brown, S. S. Venkata, and R. D. Christie, “Hybrid Reliability Optimization Methods for Electric Power Distribution Systems,” International Conference on Intelligent Systems Applications to Power Systems, Seoul, Korea, IEEE, July 1997. 40. R. E. Brown, S. Gupta, R. D. Christie, S. S. Venkata, and R. D. Fletcher, “Automated Primary Distribution System Design: Reliability and Cost Optimization,” 1996 IEEE/PES Transmission and Distribution Conference, Los Angeles, CA, Sept., 1996, pp. 1-6. 41. R. E. Brown, S. S. Gupta, R. D. Christie, and S. S. Venkata, “A Genetic Algorithm for Reliable Distribution System Design,” International Conference on Intelligent Systems Applications to Power Systems, Orlando, FL, January 1996, pp. 29-33.
18-59
(PG&E-4)
Technical Articles 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
P. Musser, R. E. Brown, T. Eyford, and C. Warren, “Too Many Routes of Reliability,” Transmission and Distribution World, June 2004, pp. 17-22. R. E. Brown and Bruce G. Humphrey, “Asset Management for Transmission and Distribution,” (to be published in IEEE Power and Energy Magazine). R. E. Brown, “Asset Management: Balancing Performance, Cost, and Risk,” EnergyPulse Special Issue on Asset Management, www.energycentral.com. T. M. Taylor, R. E. Brown, M. L. Chan, R. H. Fletcher, S. Larson, T. McDermott, and A. Pahwa, “Planning for Effective Distribution,” IEEE Power and Energy Magazine, Vol. 1, No. 5, September/October 2003, pp. 54-62. R. E. Brown and L. A. A. Freeman, “A Cost/Benefit Comparison of Reliability Improvement Strategies,” Electric Power and Light, May 2003. R. E. Brown, H. Kazemzadeh, B. R. Williams and C. B. Mansfield, “Engineering Tools Move into Cyberspace,” Transmission and Distribution World, March 2003, pp. 27-36. F. Li, L. A. A. Freeman and R. E. Brown, “Web-Enabling Applications for Outsourced Computing,” IEEE Power and Energy Magazine, Vol. 1, No. 1, January/February 2003, pp. 53-57. P. Perani and R. E. Brown, “Maintaining Reliable Power For Semiconductor Manufacture,” What’s New in Electronics, March 2002. P. Perani and R. E. Brown, “Rock Steady: The Importance of Reliable Power Distribution in Microprocessor Manufacturing Plants,” ABB Review, No. 3, 2002, pp. 29-33. H. L. Willis and R. E. Brown, “Is DG Ready for the Last Mile?” Power Quality (cover story), March 2002. pp. 16-21. R. E. Brown and M. W. Marshall, “The Cost of Reliability,” Transmission and Distribution World (cover story), Dec. 2001, pp. 13-20. R. E. Brown, P. R. Jones and S. Trotter, “Planning for Reliability,” Trans-Power Europe, Vol. 1, No. 1. March 2001, pp. 10-12. R. E. Brown, A. P. Hanson, H. L. Willis, F. A. Luedtke, M. F. Born, “Assessing the Reliability of Distribution Systems,” IEEE Computer Applications in Power, Vol. 14, No. 1, Jan. 2001, pp. 44-49. R. E. Brown and B. Howe, “Optimal Deployment of Reliability Investments,” E-Source, Power Quality Series: PQ-6, March 2000.
Invited Presentations 1.
Session Chair, “Project Evaluation and Selection,” 2004 IEEE/PES Transmission and Distribution Conference, New Orleans, LA, Oct. 2004. 2. Session Chair, “Distribution Planning and Implementation Issues for Modern Power Systems,” IEEE PES General Meeting, San Francisco, CA, June 2005. 3. Panel Member, “Assessing the impact on reliability indices after adding an OMS,” 2004 IEEE/PES Transmission and Distribution Conference, New Orleans, LA, Oct. 2004. 4. Panel Member, “Effects of System Design on Reliability,” 2004 IEEE/PES Transmission and Distribution Conference, New Orleans, LA, Oct. 2004. 5. Session Chair, “Planning Non-Traditional Distribution Systems,” IEEE Power Systems Conference and Exposition, New York, NY, Oct. 2004. 6. Speaker, “Asset Management and Financial Risk,” Conference on Probabilistic Methods Applied to Power Systems, Ames, Iowa, Sept. 2004. 7. Session Chair, “Equipment Failure Rates,” IEEE PES General Meeting, Denver, CO, June 2004. 8. Speaker, “Distribution Asset Management,” 2003 Real World Conference: It’s All About Cost and Reliability, Transmission and Distribution World, Ft. Lauderdale, FL, Oct. 2003. 9. Speaker, “The 2004 Northeast Blackout,” NC State IEEE/PES Student Chapter, Oct. 2003. 10. Panel Member, “Distribution Reliability Standards and Their Basis,” 2003 IEEE/PES Transmission and Distribution Conference, Dallas, TX, Sept. 2003. 11. Session Chair, “Power System Planning in an Evolving Regulatory Environment,” IEEE PES Summer Power Meeting, Toronto, Ontario, July 2003.
18-60
(PG&E-4) 12. Panel Member, “Distribution System Reconfiguration,” IEEE PES Summer Power Meeting, Toronto, Ontario, July 2003. 13. Panel Member, “IT Solutions For Distribution System Planning,” IEEE PES Summer Power Meeting, Chicago, IL, July 2002. 14. Panel Member, “Distribution System Reliability Assessment,” IEEE PES Summer Power Meeting, Chicago, IL, July 2002. 15. Panel Member, “Current Status of DG Models for Feeder Analysis,” IEEE PES Summer Power Meeting, Chicago, IL, July 2002. 16. Speaker, “Tools for Cost-Effectively Improving Reliability,” Managing Distribution Systems in a Deregulated Environment, EUCI, Denver, CO., May. 2002. 17. Session Chair, “Impact of DG on System Reliability,” Power Systems 2002 Conference: Impact of Distributed Generation, Clemson, SC, March 2002. 18. Speaker, “How to Apply Reliability Improvement Methods to Your Distribution System,” Electric Distribution Reliability Planning Conference, INFOCAST, Seattle, WA, Nov. 2001. 19. Panel Member, “Status of Distribution Reliability in the United States,” 2001 IEEE/PES Transmission and Distribution Conference, Atlanta, GA, Oct. 2001. 20. Panel Member, “Distribution System Reliability and Reconfiguration Software Tools,” 2001 IEEE/PES Transmission and Distribution Conference, Atlanta, GA, Oct. 2001. 21. Panel Member, “Challenges in Distribution System Analysis,” IEEE PES Summer Power Meeting, Vancouver, BC, Canada, July 2001. 22. Panel Member, “What are the Appropriate Reliability Targets for Distribution Companies to Meet?” IEEE PES Winter Power Meeting, Columbus, OH, Jan. 2001. 23. Speaker, “Distribution Reliability Challenges,” Distribution System Planning, Maintenance and Reliability Conference, EUCI, Denver, CO., Nov. 2000. 24. Speaker, “Reliability-Based Planning Methods: How to Choose a Method That Best Meets Your Reliability Goals,” Electric Distribution Reliability Planning Conference, INFOCAST, Chicago, IL, Sept. 2000. 25. Speaker, “The Impact of Deregulation on Electric Power System Reliability,” CUEPRA Summer Meeting, Charlotte, NC, July 2000. 26. Speaker, “Tools for Analyzing and Valuing Distribution Reliability,” Power Delivery Reliability Conference, INFOCAST, Denver, CO, June 2000. 27. Panel Member, “Rates and Reliability—Peaceful Co-Existence,” DistribuTECH Conference, Miami, FL, Feb. 2000. 28. Speaker, “Optimizing Distribution Reliability at Minimum cost Using Computer Optimization,” Improving Distribution Reliability Conference, Washington D.C., Jan. 2000. 29. Speaker, “Managing Cost, Reliability, and Financial Risk for Power Distribution Systems”, E-Source Power Quality Summit, Chicago, IL, Nov. 1999. 30. Speaker, “Noteworthy Topics in Power System Planning in a Deregulated Environment”, IEEE PES Winter Power Meeting, New York, NY, Feb. 1999. 31. Speaker, “Distribution Reliability for De-Regulated Utilities”, IEEE PES Winter Power Meeting, New York, NY, Feb. 1999. 32. Speaker, “Design for Reliability: What Level of Reliability Should Distribution Systems Be Built For?” Rethinking Electricity Distribution Reliability Conference, INFOCCAST, Atlanta, GA, March 1998. 33. Panel Member, “Value of Reliability for Distribution Systems,” DistribuTECH Conference, Tampa, FL, Jan. 1998. 34. Speaker, “Design for Reliability: What Level of Reliability Should Distribution Systems Be Built For?” Rethinking Electricity Distribution Reliability Conference, INFOCAST, Chicago, IL, Sept. 1997. 35. Speaker, “Distribution System Design: Reliability and Cost Optimization,” IEEE/PES Seattle Section, Seattle, WA, May 1996. 36. Speaker, “Power System Reliability Assessment,” University of Washington Electric Energy Systems Seminar, Seattle, WA, Sept. 1995.
18-61
