ML20246J381

Corrected Rev 1 to AEOD/S901, Maint Problems at Nuclear Power Plants

ML20246J381
Person / Time
Issue date:	05/31/1989
From:	Chiramal M, Israel S, Wegner M NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD)
To:
Shared Package
ML19302D909	List: ML20246J381 ML20246J442
References
TASK-AE, TASK-S901 AEOD-S901, NUDOCS 8905170064
Download: ML20246J381 (49)

Text

.- . _ - - - . - - -

,; t -

y :Q, y 9; .

AE0D/S901, Rev. I m

SPECIAL STUDY REPORT AEOD/S901, REVISION 1 11AINTENANCE PROBLEMS AT NUCLEAR POWER PLANTS Mey 1989 Prepared by:

Mary Wegner Matthew Chiramal Sandy Israel' Sal Salah Stephen Stern 8905170064 890509 PDR ORG NEXD PNU 4

I Revisions are shown with a colon (:) in the right hand margin of the page.

r

.t

't l

SUMMARY

A measure of operating nuclear plant maintenance deficiencies was obtained by reviewing operational experience reported to the NRC over the last 4 years. The review focused on existing studies and reports by the Office for Analysis and Evaluation of Operational Data, generic communications by the NRC to licensees, feedback documents issued by the Institute of Nuclear Power Operations (INPO),

and Licensee Event Reports issued by operating domestic nuclear plants. These documents were reviewed to identify maintenance-related problems and deficien-

, cies, the number of plants involved, the systems and components affected, and l the human and organizational errors associated with maintenance. The review attempted to ascertain if there has been a discernible trend in the area of maintenance. The study also addressed estimates of the magnitude and trends in the cost of identified maintenance deficiencies to the commercial nuclear industry.

Of the 70 AE0D case studies, special studies, and engineering evaluations issued since 1985, the review identified 15 (about 20% of all studies) that specifically addressed maintenance-related problems. We conclude that the maintenance-related problems found in systems such as the service water system, the instrument air system, and other safety- and non-safety- related systems, and in components like motor-operated valves, inverters, circuit breakers, and pumps, are widespread in the industry. Problems reflect deficiencies in quality control, procedures, planning, communication, and training.

Since 1985 the NRC has issued 80 bulletins and information notices specifically addressing maintenance deficiencies (approximately 20% of a total of about 400). The review of these documents provide similar conclusions to those obtained by the review of the study reports - i.e., problems addressed in these generic communications identified maintenance procedure deficiencies, inade-quate post-maintenance testing, maintenance problems of motor-operated valves, emergency diesel generators, pumps, valves, relays, etc. Similar findings were identified in the review of generic feedback documents issued by INp0.

These documents, i.e., AE00 reports, and NRC bulletins and information  :

notices, characterize the broad scope of maintenance deficiencies and the  :

significant staff effort to alert licensees of maintenance problems. Some of:

these documents are based on operating experience over several years, while  :

others are based on one or more safety significant events. Simple counts of :

the documents cannot be meaningfully used to establish trends. In order to  :

ascertain trends in reported maintenanate deficiencies, we relied on data from:

operating reactors reported in licensee event reports (LERs) contained in  :

the Secuence Coding and Search System (SCSS). This effort identified a  :

gradually improving trend in maintenance-related problems in the industry in :

the 4-year period beginning with 1985. It should be noted that LERs focus on :

safety-related equipment and systems, and hence, trends based upon LERS are  :

primarily associated with such equipment and systems.  :

Improving trends in forced outage rate and equipment forced outages, and a  :

marked improvement in the scram rate, all support the view that overall plant :

maintenance is improving.

Hence, we conclude that although maintenance problems continue to occur in a :

broad number of systems and equipment, recent operating experience point to an:

improving trend in safety performance that can be attributed to improvements :

in maintenance.  : ;

A bounded estimate for the cost of maintenance deficiencies associated with replacement power costs for forced outages was developed from the SINET data base. The upper bound of this estimate shows steadily increasing replacement i power costs from 1984 to 1987, reaching an annual level of 960 millicn dollars in 1987. The upper bound for 1988 is currently projected at 720 million dollars per year which is a decrease from 1987. The lower-bound estimates are 120 million dollars in 1987 and a projected 80 million dollars in 1988.

This report can provide additional insights for ongoing maintenance team inspections, as well as provide some basis to consider modifications to the inspection modules. Additionally, the results can provide useful data to further an understanding of the quality of maintenance at operating plants, and serve as input to a regulatory guide on maintenance. Lastly, the information regarding financial matters can provide additional perspective regarding current maintenance costs.

1

l

. . J

.t

1. INTRODUCTION This stuoy was initiated to determine the extent and pervasiveness of maintenance problems in operating nuclear power plants. Maintenance is defined as the aggregate of those actions required to preserve and promptly restore  !

operability, reliability, and availability of, or to prevent the failure of )

plant structures, systems, and components.

The study is based on operational experience reported to the NRC, and is based on the review of existing completed AEOD studies, NRC generic communica-tions to licensees, and licensee event reports. The review of these documents focuses on maintenance related problems and deficiencies.

TI,e review identifies plant designs that have been affected by maintenance related problems and deficiencies, the systems involved, and the components that predominate. Finally, the review attempts to identify any trends in the area of maintenance in the recent years.

AE00 studies based on operating experience are discussed in Section 2.1, Maintenance related NRC bulletins and information notices are reviewed in Section 2.2, and similar INP0 documents in Section 2.3. Trends based on available licensee event report data bases are discussed in Section 2.4 and 2.5. Section 2.6 discusses NRC Performance Indicator program data related to maintenance activities. Section 2.7 discusses the economic assessment of maintenance deficiencies.

2. OPERATIONAL EXPERIENCE REVIEW 2.1. Review of AE00 Reports of Case Studies, Special Studies, and Engineering Evaluations Issued since 1985 AEOD has issued 70 reports since 1985 which discuss important safety concerns brought to light by reactor operating experience. Licensee event reports and

! other related documents are screened to identify significant issues which l warrant further analysis by the staff to determine the salient factors involved in the issue including the underlying causes and the extent of industry exposure. Once a suitable topic has been identified, all reports of similar experience are amassed to provide a knowledge base for the AE00 study.

l Reports issued by AE0D since 1985 were screened to identify those issues where l the quality of maintenance was directly related tosthe issue. The reports l selected for review are:

1. C801 - Service Water System Failures and Degradations in Light Water Reactors.

2. 5801 - Significant Events Involving Procedures.

3. NUREG-1275, Vol. 2 - Operating Experience Feedback Report - Air Systems Problems.

4 E702 - NOV Failure due to Hydraulic Lockup from Excessive Grease in Spring Pack. .c

5. C603 - A Review of Motor-Operated Valve Performance.

6. S503 - Evaluation of Recent Valve Operator Motor Burnout Events.

. . 1

(

^ '

~2-

.s

7. C502 - Overpressurization of Emergency Core Cooling Systems in l BWRs.

8. C504 - Loss of Safety System Function Events.

9. P701 - Trends and Patterns Program Repcrt - Operational Experience Feedback on Main Feedwater Flow Control and Main Feedwater Flow Bypass Valves and Valve Operators.

10. E802 - Single Failure and Other Deficiencies Noted in Control Room Emergency Ventilation Systems.

11. E804 - Reliability of Recirculation Pump Trip Breaker During an ATWS.

12. E708 - Depressurization of Reactor Coolant Systems in PWRs.

13. C605 - Operational Experience Involving Losses of Electrical Inverters.

14. E504 - Loss or Actuation of Various Safety-Related Equipment Due to Removal of Fuses or Opening of Circuit Breakers.

15. E512 - Failure of Safety-Related Pumps Due to Debris.

The details of the review are discussed as follows:

" Service Water System Failures and Degradations in Light Water Reactors" (AE0D/C801) is a case study that compiled almost 1000 reactor events related to the subject from 1980 to 1987. About 25 percent of the events were judged to have generic safety significance. Systems and components adversely affected by a service water system failure or degradation include the component cooling water system, emergency diesel generators, emergency core cooling system pumps and heat exchangers, the residual heat removal system, containment spray ar.d fan coolers, control room chillers, and reactor building cooling units. Risk estimates indicate that the safety significance of service water system failures and degradations is relatively high.

System fouling caused a majority (58%) of the 276 events thet had potential generic significance. This degradation, which is a chronic maintenance problem, includes sediment deposition, biofouling, corrosion / erosion, pipe coating failure / carbonate deposition, and foreign material / debris intrusion.

Although modifications were made at some plants to correct some of the problems in this area, the major long term solution appears to be effective surveillance and preventive maintenance practices. Personnel and procedural errors related to maintenance activities accounted for another 13 percent of the major events.

These events include leaving the system misaligned af ter maintenance, improperly installing components, and inadequate coordination with operational requirements. These situations reflect problems with quality control, procedures, and communications.

The events caused by maintenance related deficiencies cited in the report occurred at about half of the plants. Thus, these problems are wide spread.

Since the service water system is a direct conduit to the plant heat sink, degradation of the capacity of the heat exchangers in the various safef.< .

systems supported by service water is important. None of the plant Technical Specifications reviewed, however, requires licensees to perform any type of heat exchanger testing. It is important to note that severa1>affected plants have instituted comprehensive surveillance programs to monitdr service water I degradation in selected components. At on plant the tests are performed l

l l

. i 5

weekly while at another they are performed monthly. These frequencies are higher than some of the required tests on safety systems and attest to the importance of the concern and severity of the problem.

" Operating Experience Feedback Report - Air Systems Problems" (NUREG-1275 v.2) compiled about 160 events involving failure of safety related equipment caused by degraded or malfunctioning non-safety grade air systems. Approximately 40 percent of these events are concerned with deficiencies in the maintenance of air- actuated components or the air system itself. The safety systems affected by the air-system events include the reactor trip system, auxiliary feedwater, residual heat removal, emergency core cooling system, containment isolation, main steam isolation vulves, control room ventilation, waste gas treatment, and emergency electric power. Because of the potential for common cause failures, these problems represent a significant safety concern.

A principal maintenance deficiency pointed out by this study is the lack of effective monitoring and maintenance of the air system which resulted in water and other contaminants building up in the system and failing air actuated components. In this regard, the air solenoid valves appear to be particularly susceptible to contaminants because of small clearances. The study finds that maintenance personnel frequently did not understand the potential safety consequences of degraded air systems. They are often unaware of the potential for common cause failure of redundant safety related equipment that rely upon air systems. Significant events which occurred at 31 plants are documented in the study and cover older plants as well as recently licensed plants.

Motor operated valve (MOV) problems were reported in a number of studies that find maintenance a principal contributor to equipment degradation and failure.

Most recently, motor operated valves received considerable attention because of chronic and pervasive problems. The following reports were issued on this topic:

MOV Failure Due to Hydraulic Lockup from Excessive Grease in Spring Pack (AE00/E702)

A Review of Motor-operated Valve Performance (AE0D/C603)

Evaluation of Recent Valve Operator Motor Burnout Events (AE0D/S503)

Since uotor operated valves exist in most fluid systems in the plant, the problems are widespread. Five hundred and sixty five LERs involved problems with MOVs in the time period 1981 through mid-1985. The NPRDS data bae, which is managed by the Institute of Nuclear Power Operations, yielded just over 1100 entries in the same time period. The data have not been specifically analyzed i to determine which events were strictly maintenance related; however, the '

evidence suggests that between 50 and 80 percent of the events fall into this ca tegory. Because flawed maintenance practices involving MOVs pervade all systems, the potential for common cause failure of redundant safety equipment is significant.

Improper torque / limit switch settings is a major contributor'to degraded valve operation. It is hypothesized that torque switch " corrective action" may have substituted for definitive identification of the underlying valve inoperability

4 problem. Thus, it is conjectures that the flexibility to make adjustments rather than perform a root cause analysis may have bred bad maintenance habits.

One of the studies addressed the use of a new environmentally cualified grease which had a lower viscosity. The lower viscosity grease could migrate more easily to an area where a spring must compress to provide thrust to operate the POV. Due to inadequate greese relief in the spring area, the spring could not compress. Thus, inadequate design evaluation resulted in addressing one problem (equipment qualification) while creating another problem (spring pack or hyoraulic lockup). MOV events have occurred at virtually every plant.

"Overpressurization of Emergency Core Cooling Systems in Boiling Water Reactors" (AE00/C502) is a case study of potential intersystem loss-of-coolant-accidents caused overpressurization of low pressure sytems due to the pressure boundary interface problems. Eight events are cited which involve a stuck open isolation check valve at the pressure-boundary. Five of the these events were caused by maintenance errors. The safety significance of this issue is the potential for an unisolable LOCA outside containment. The maintenance errors include equipment misalignment and inadequate preventive maintenance. The report cites occurrences at four of the 26 BWR sites.

" Loss of Safety System Function Events" (AE00/C504) is a case study that compiled 87 events of loss of safety system function involving human factors in the time period 1981 to 1984. Approximately 28 percent of these events occurred during maintenance activities. Most of the events cited involved the residual heat removal system or the ECCS. The types of human errors that occurred during maintenance include failure to follow procedures, inadequate planning, defective procedures, and failure to perform work correctly. These events occurred at 20 different plant sites.

" Trends and Patterns Program Report - Operational Experience Feedback on Main Feedwater Flow Control and Main Feedwater Flow Bypass Valves and Valve Operators" (AE00/P701) provides an engineering analysis of reported failures in these components. The report finds that the main source of variation in failure rate is due to the differences in maintenance practices among units and stations. The evaluation shows that proper maintenance and the use of appro-priate subcomponeats are dominant positive actions to avoid problems in the main feedwater system. Valve operator failures were caused by poorly maintained air systems which allowed water, oil, and foreign particles to enter the system.

These types of problems are discussed aleve in the air system report. Poor maintenance procedures in some plants caused repeated failures. Detailed procedures, use of routine maintenance schedules, .and assistance of manufac-turer's representatives were used to correct the problems of these valves.

" Single Failures and Other Deficiencies Noted in Control Room Emergency Ventilation Systems" (AEOD/E802) is a review of nine events involving an inadequate emergency ventilation system. Four of the events were caused by maintenance errors - misaligned systems or reversed leads. A failed emergency ventilation system could imperil the operating staff and impair the response to certain accidents such as a toxic gas release. The events occurred at four ,

different plants. ,j

" Reliability of Recirculation Pump Trip Breaker during an ATWS" (AE0D/E804) compiled 22 breaker failures for the time frame 1979 to 1987. According to the manufacturer, most of the failures to trip on demand were caused by circuit

breaker misadjustment or lubrication problems. The frequency of reports appears to have decreased subsequent to the manufacturer issuing a service information letter to the utilities. This problem was reported at 8 of the 26 BWR plants.

"Depressurization of Reactor Coolant Systems in PWRs" (AE0D/E708) examined the impact of selected systems that were out of service during a loss of power event at Salem in 1986. Several components that could control system pressure were out of service for extended periods of time prior to the Salem event because corrective maintenance was deferred until the next refueling (expected to be several weeks hence). In addition, several other pieces of non-safety grade equipment were also inoperative because of deferred maintenance. This situation could have distracted an operator, responding to a more challenging event, because of missing process information and faulty equipment that would require doditional operator attention. Much of the equipment were not in the limiting conditions for operation in the plant Technical Specifications and thus were permitted to be disabled for an extended time. Similar deferred maintenance activities were noted at other plants.

" Operational Experience Involving Losses of Electrical Inverters" (AE0D/C605) examines 142 events for 1982 to 1984 Personnel actions were responsible for 35 percent of these losses. These types of losses have significant safety impact because they represent the degradation of the vital instrument power supply and potentially the common cause loss of the entire instrumentation sys tem. The types of personnel activity that have affected the inverters include opening incorrect circuit breakers, connecting faulty test equipment to the inverter, causing short circuits during maintenance or testing, and removing the wrong unit from service. The observed problems suggest that improvements are needed in planning, procedures, verification methods, and training practices regarding testing and maintenance activities. These events occurred at 25 plants.

" Loss or Actuation of Various Safety Related Equipment Due to Removal of Fuses or Opening of Circuit Breakers" (AE0D/E504) discusses six events which occurred in 1981 to 1984. All of these events are related to system misalignment which occurred during testing or maintenance activities. The report asserts that the number of events are understated because of the limited reporting requirements.

This type of personnel error leaves the equipment in a failed state that will not be caught until the component is challenged or effective post maintenance verification is performed.

" Failure of Safety-Related Pumps due to Debris" (AE0D/E512) is a report on potential common cause failures from sloppy maintenance activities. The study cites five events where debris caused component failure and three other events l that flagged poor housekeeping inside of containment following maintenance activities.

Roughly 20 per cent of the reports issued by AE00 in this time period identified issues where maintenance was a significant causative factor. These studies showed that maintenance problems and deficiencies occ,urred at a large number of plants and affect a large number of plant systems. More importantly, I these studies are judged to have generic safety significance. l l

l l

The maintenance problems vary among the studies reviewed. They include inadequate preventive maintenance as illustrated by the chronic problems with

) the service water system and the air systems, improper maintenance as shown by l

the problems with the motor operated >>ives and the electrical inverters, and system misalignment after maintenance as discussed in the reports on pressure boundary configurations in BWRs and control ventilation systems.

2.2. Review of NRC Bulletins and Information Notices Issued Since 1985 Review and analysis of NRC generic communication documents, such as bulletins and information notices, is another approach to identifying the maintenance-related problems that are being experienced in operating nuclear plants. These documents alert the industry of problems and the need for corrective actions on specific items that are identified by operatior.al experience at one or more nuclear plants. All NRC bulletins and information notices issued since 1985 were screened to select those involving a maintenance related issue. During this period a total of 23 bulletins and 369 information notices were issued -

in 1985 four bulletins and 101 information notices, in 1986 four bulletins and 123 information notices, in 1987 two bulletins and 69 information notices, and by November 1988 12 bulletins and 86 information notices. The screening effort identified 27 maintenance related bulletins and information notices in 1985, 36 in 1986, 12 in 1987, and 10 in 1988. These documents were reviewed for the maintenance related problems involved, the system or component affected and the plants involved. The details of this review are shown in Table 1.

NRC information notices and bulletins generally discuss industry-wide problems which are of sufficient safety significance and concern to merit both NRC and industry attention. They are issued to all operating nuclear plants for information and action, as appropriate. Hence, the maintenance problems that are addressed by the documents in Table 1, are problems that are considered generic and potentially applicable to many, and perhaps all plants.

Of the approximately 80 documents listed in Table 1, 10 dealt with maintenance related problems of motor-operated valves, 10 involved maintenance procedure.

deficiencies, 5 addressed inadequate post-maintenance problems, 5 each related to maintenance related problems of the control rod drive systems and the emergency diesel generator system of PWRs and BWRs. 4 each involved inadequate inservice testing, maintenance problems of safety valves, and human errors during maintenance activities, and 3 each addressed maintenance related problems of steam generators and containment systems. The remainder of the notices dealt with maintenance problems of variousssystems and components, such as auxiliary feedwater pumps, plant air systems, instrument transmitters, relays, de batteries, circuit breakers, relief valves, solenoid valves, pump seals, and service water systems. The maintenance related problems of components and equipment such as motor-operated valves, control rod drives, emergency diesel generators, safety valves, pumps, circuit breakers, etc., can be ,icintly considered as deficiencies in the plant preventive maintenance and housekeeping programs.

2.3. Review of Industry Reports ,j Tha INP0 reports, like the NRC bulletins and information notices, are nuclear plant operational experience feedback documents that address industry-wide problems that are deemed sufficiently significant by INPO and the industry to

merit attention. Although the maintenance related problems addressed by these documents occurred at specific plant sites, the problems are potentialll/ i generic to all operating nuclear plants.

All INPO significant operating experience reports (SOERs), significant event reports (SERs), and operation and maintenance reports (0&MRs) issued in 1987 and 1988 were screened to identify those related to a maintenance problem. The 3 SOERs, 24 SERs, and 18 0&MRs thus identified were reviewed for the type of problem involved, the system or component affected, and the plants where the event occurred. The details of this review are shown in Table 2.

Of the 45 reports in Table 2,14 involved inadequate planning and procedures, 12 pertained to preventive maintenance deficiencies, 10 concerned human errors committed, and 5 involved inadequate testing of equipment and systems. The remaining reports dealt with various equipment and component problems.

2.4. Licensee Event Report Database Review  :

Analysis of databases that capture various reports on nuclear power plant  :

experience was done to estimate the trend of maintenance-related activities in:

the industry. We relied heavily on LERs encoded in the Sequence Coding and  :

Search System (SCSS). We also used data from the NRC Performance Indicator :

(PI) program and from the licensee monthly operating data reports. (NOTE:  :

The original report included information from the Nuclear Document System  :

(NUDOCS), the NRC's automated, searchable document system that contains all  :

documents generated by the regulatory process. Trending of maintenance-related documents contained in NUDOCS is a measure of maintenance deficiencies at  :

nuclear plants as well as a measure of NRC staff's focus on maintenance issues.

Since we could not readily sort these aspects of the data, we deleted the  :

NUDOCS data from this revised report.)  :

SCSS is a detailed data base of encoded LERs which includes information about systems and components challenged during an event and the transient aspects of the event. The system primarily captures safety-related deficiencies. While :

balance-of-plant (B0P) initiated trips are reported, events involving B0P  :

equipment may not need to be reported by licensees. Nevertheless, since the :

SCSS is a data base system that undergoes rigorous quality assurance checks,  :

its consistency lends itself to trending (time dependent) analyses.  :

Initially a narrow search for maintenance deficiencies was performed using the:

SCSS data base with the codes "AB", "AU", and "AW"<, "AB" refers to a component or part that is present, but is not in its proper location due to improper adjustment or positioning. This does not include improper setting or miscalibrated instrument. It does include valve installed backwards. "AU" refers to an improper previous repair while, "AW" describes a previous repair that was done correctly but was ineffective at preventing recurrence of the component failure. We found 120 records in 1985, 135 in 1986, 160 in 1987, :

l and 108 in 1988. The trend here shows an increase till 1987, and a decrease :

l in 1988. However, it should be noted that the data average about one per  :

plant per year, and hence, there is insufficient data to support an indication:

of industry trend.  :

Broad searches involving thousands of records and using large number of codes :

- codes in the order of 20s - provide more meaningful trends. Such an effort : 1 is described in the remainder of this section.  :

i I

8-t -

LERs in the SCSS were used in a program by the Oak Ridge National Laboiatory  :

(ORNL) to extract cause and corrective action data.* This cause data is  :

distributed among the following seven cause codes: (1) administrative control:

(AC), (2) random equipment failures (REF), (3) design, fabrication,  :

construction, and installation error (DFCI), (4) licensed operator error (LOE),

-(5 other personnel error (OPE), (6) maintenance 1 error (MI), and  :

(7 maintenance 2 error (M2).  :

The cause code data are extracted as numbers of counts in each of the seven :

cause codes. An LER containing more than one of the seven cause codes is  :

counted in each cause code category. An LER containing multiple counts of the:

same cause code is counted only once for that code. On an average, about two :

causal factors (counts) are associated with each LER.  :

Twomaintenance-relatedcaLsecodesareused-maintenance 1(M)1and  :

maintenance 2 (M2). M1 and M2 are mutually exclusive and, hence, may be  :

summed. Both measures are expansive and, in total, are intended to capture  :

all aspects of a maintenance program. M1 captures those deficiencies clearly :

related to implementation of a maintenance program. M2 captures equipment  :

failures that may be indicative of a maintenance program deficiency.  :

The maintenance 1 error cause code intends to reflect deficiencies that are  :

attributable to poor maintenance practices or maintenance personnel errors.  :

Data under this code is obtained by finding and counting, avoiding double  :

counting, (1) all LERs coded in the SCSS as personnel activity errors in the  :

maintenance, surveillance, radiation protection, or " unknown" category, and  :

(2) all LERs with one of the following causes: missing, mispositioned,  :

improper separation, loose, tight, wrong part, wrong material, improper  :

repair, ineffective repair, impact, dropped, worn, boron precipitation,  :

corrosion, erosion, open circuit, short circuit, drift, aquatic organism,  :

weld-related flaw, foreign object, or lubrication-related problem. Of these :

LERs only those in which repair or adjustment is required is counted as an M1 :

category event.  :

The maintenance 2 error cause code reflects problems which are potentially  :

indicative of a maintenance problem. It is clearly intended to capture events:

beyond those identified by M1. Data for this ccde is obtained by first  :

finding all LERs in the SCSS encoded as: temporary modification, overload,  :

fatigue, vibration, air / steam binding, cavitation, loss of pump suction,  :

pressure pulse, high temperature, freezing condition, high thermal change  :

rate, overcurrent, undercurrent, overvoltage, undervoltage, instrument  :

repeatability problem, electrical noise, water spray, high humidity, smoke, :

radiation, high ambient temperature, cladding degradation, and various "other":

codes. Also included are " unknown" causes for non-solid state type  :

components. Failures attributable to an environmental factor, such as  :

lightning, are excluded. Events which were linked back to a previous  :

personnel error and events which involve postulated or potential failures are :

• 0RNL/NOAC-244, Results from Trial Program for Use of Cause and Corrective Action Data as Programmatic Performance Indicators. January 13, 1989, Oak Ridge National Laboratory /NRC.

L

<- removed from the data. Then, from the remaining, only those events in which  :

repair or adjustment is required are counted as M2.  :

In the original report,.the cause code counts for three 12-month periods  :

(1) July 1, 1985 - June 30, 1986, (2) July 1, 1986 - June 30, 1987, and (3) :

July 1,11987 - June 30, 1988_, were displayed to show the distribution of the :

counts as pie charts with percentages for.each cause code. The total number :

of LERs for each 12-month period was also given. In that report the three  :

charts indicated that the distribution of maintenance 1 and maintenance 2  :

errors remained rather constant (M1 ranged between 31 and 32 percent, and M2  :

while the number of LERs over the three periods  :

between11and15 percent)(from2815to2550).

showed a cecreasing trend  :

For this revision, the data for the seven cause. codes are expanded to cover :

all four quarters of 1985, 1986, and 1987, and the first two quarters of 1988.:

The distribution of the cause code count data as a function of time in years :

is shown in Figure 1A and IB. The same data is shown in Figures 2A, 28, 2C, :

and 2D as pie charts for the three years 1985, 1986, 1987, and for two-  :

quarters of 1988. (In Figures IA and IB, the data for 1988 is obtained by  :

projecting the data of the first two quarters of 1988.) These figures are for:

the industry as a whole and have not been normalized for the growth in the  :

number of operating plants, nor have the data from plants in extended  :

shutdown been excluded. As seen in these figures, the number of LERs have  :

remained generally constant (around 2750 to 2800) during 1985,1986 and 1987, :

but the number of cause code counts show an increasing pattern till 1988. The:

projected numbers for 1988 show a decrease in both the number of LERs and the :

number of cause code counts.

The pie charts in Figures 2A,2B, 20, and 20 which depict the distribution of  :

the seven cause codes for 1985, 1986, 1987, and the first two quarters of  :

1988, show that M1 has remained around 31 percent, and M2 has decreased from  :

16 percent to 11 percent. These figures (Figures 1A, 18, 2A, 28, 2C, and 2D) :

indicate that the fraction of maintenance error M1 has remained relatively  :

unchanged and the fraction of maintenance error M2 has shown a decline.  :

Overall one may conclude that maintenance problems have continued to represent:

a substantive fraction of all problems reported in LERs.  :

It should be noted that the number of plants licensed to operate have  :

increased since 1985 - 95 in 1985, 104 in 1986 and 109 in 1987 and 1988.  :

Taking this into consideration, the maintenance related cause code data M1 and:

M2 for the four quarters of 1986 and 1987, and for the two quarters of 1988, :

are normalized and an industry average maintenance indicator trend for M1 and :

M2 is obtained as shown in Figures 3A and 38. Figure 3A shows the data from :

all operating nuclear plants while Figure 3B shows data for plants licensed :

prior to 1984. The graph of M1 shows a declining trend of this indicator of  :

about 6 percent per year for all plants and 4 percent per year for mature  :

plants over the 21/2-year span. Indicator M2 shows a sharper decline of  :

about 18 percent per year for both sets of plants. These indicator trends  :

show that nuclear plant maintenance programs for safety-related systems,  :

a a t c q e N M M O iL E D d

A E

G E

L S

S C

S -

g G

N i

I -

S U

)

Sd ze l

A Ti l E '

1 Nam 1

1 E 1' RUr U

G o E FO n I 1 Cto n

(

E D -

O C

E S

U A

C

~ - -

_ ~ -

0 0 0 0 0

_ 0 0 0 0 0

_ 0 0 0 0 0 6 0 4 3 2 1 3= ~

l

_ 11 _

. +

,E b " U d 3

.E Sd O t J J I o E

'O ro O

'5

,U 2

-s *

.E O

.E O

i' Z < a CT tg _; o y 3 m, W )

i 3

e m  :

w i o i a

Nf!!N! smaer m f Z n 8 -

R w kwwwwwxv a 3 nummmmmew%%wansw -

v m m8 -ummim llllll giggll

-H i ll n in i un i nunini i i un!n in inn i N

uZ E M8MM eJg m $

8oc kwwwwxwxv ,

Eo; h% - W W $

uS I Q

O M

IHil yIIllli Illi !I IWllIBMll 1111lll 11H! H!llh' 'l musea O @

Wed ,-

y NXwwwwww w EmN%msu mu%%% m msw D

o II llllIllllllllggglIllllllllg g

-- g rm ,_

LNwwwww LWhmN\W M4hh%%%Ni l l l l l l l l l 8 8 8 8 8 8 8 8 8 8 R  % 9 s;unog jo;o1

S S

C 1 S

G t N i n

I S a U M N  %

O 9 .

I 2

AT5 1 2

U8 9 3 EB1 R t n

I U

G RR T FO i

I F a S

I M

D E

D O  %

C 5 .

E  % 4 S 5 4 0 9  %

U 0 3 8 5 6 9 p A 2 . 1 4 . 0 O C s t

s 2 1 R n .

E u s c L o C

i n p n i f

of i i g L o

r u i s

e r b e m q e mb d E .

u m N u A D w

I I  ! l 1 l l il 1

~

S S *  %

C 1 4 _

S 3 t l 1 -

G n -

N i 2 -

I a

S M U t

% n N

i .

O 1 a -

M I

1 BT6 2

U89 3 6 EB1 R I 8

U 9 G RR O 1 F TF I

S I

D E 9  %

1 D .

O 6 _.

C . -

p _

E s  %

S ~

7 i

n 2 O _

U 8

4 5 1

i  % . _

A 7 5 3 3 c C

2

m . 1 i st s d 2 L R n E u A .

n L o g f

C p i of r

o i

u s e r b e q e mb um E D N u N

_ l

7 .

5 1

P S r S e  %

C 1 h 6 .

S t 1 t

O 1 G n N

I i

a 2 S M U t

% n N 8 i

a O .

M I

CT7 0 2

U89 3 EB1 R I U

G RRO FI TF S

I D  %

4 E 1 8 .

D 2 5 O

C p E .  % O S 5 i s 9 U 2 0 6 9 n 2 c A 7 5 2

i 8 1 i C .

s m 1 L

s R n t

d n E u L o A . g C p i f

of i s r

e r o u q

e b e mb D

um E '

N u N

P S r 7 S e .

C h 0 t

S 1 O

G 2 N

I t

S  % n U 4 i

)

S  % 1 a NRE 1 M OTR .

TU A I

1 2UQ D 3 EBO R W UI G R (T I

F T8 8 S9 I 1 DRO  %  % 7 .

F 2 4 5 E .

1 D 3 .

O 2 p C n O g

E i .

S s s c U

i

% e i L

A 3 8 1

6 8 i n 3 D 2 2 .

C 1 m 1 s ts R n E u d .

L f

o f C

o A .i p

o r

e r e

u b

mb q um N u n

E .

E G o a E

L D -

N O X .

w _

E R

T O--

X RST ON - -

TLA -

AP O -

CLL ADA I

3 N R)

P X I

Od E Fz e -

RE i .

l Ea - .

U SGm U Ao r - -

G Rn U' -

x .o I

A VE( -

FCA -

Y ER - -

X .

+

CTS .

NUD .

A N o NI -

. O E

T - .

I N

A -

M D - X X ,

e -

7 6 5 4 j n

,W l  !

l

E G x a E

L D48 -

N9 1 a- X -

E RO T T R '

RIRO OP _

T _ -

ADE " X O CS I

N . - _ -

8DE C .

3NI L )d I X .

e _

E ETl Szi a .

R S Nm Ar _

U ULo a X G A Pn

(

FCRO I

F E

E a X _

CGA

~

~

N R .

AE -

NAV

. o O .

E Y TR NTS -

I a X g A DU MN I O .

X

. X .

o _

6 ca 4 3 2 3 N y$ i

.. i l

equipment, and components are effecting a decrease in the number of reported  :

events involving failures of such systems, equipment, and components.  :

The overall conclusion reached in reviewing the various aspects of the  :

maintenance error cause codes extracted from LERs is that in the past 3 years :

maintenance-related problems experienced by operating nuclear plants have been:

decreasing. )

2.5 Review of Events Identified as Maintenance-Related Safety System Failures In Section 2.4 data based on all LERs was presented. In this section, we  :

focus on that subset of LERs used in the NRC Performance Indicator (PI)  :

program. The PI program classifies and tracks those LERs which involve  : I automatic scrams (scrams), safety system failures (SSFs), safety system  :

actuations (SSAs), and significant events (SEs). These PI LERs are considered:

to be of a higher safety significance by the NRC than other LERs.  :

1 By trending those LERs which are in the PI program and which can be  :

attributable to maintenance deficiencies as identified by the SCSS cause code :

extraction scheme (M1 and N2), an estimate of the trend in significant  :

maintenance related events can be obtained. Such a trend for three quarters :

of data was included in the original report. For this revision, the trend  :

for six quarters, beginning the first quarter of 1987 and ending the second  :

quarter of 1988, is obtained.  :

Figure 4.1A and Figure 4.1B show the distribution of events in the PI programs:

which were classified as describing an M1 error . Figures 4.2A and 4.28 show :

the data distribution for M2 events. The figures show a decrease in the  :

fourth quarter of 1987 and followed by a small decrease in the two quarters of:

1988. As seen in Figures 4.1B and 4.28, the decrease is mainly due to the  :

drop in the number of scrams and safety system failures starting in the fourth:

quarter of 1987. The trends can be interpreted to mean that the industry  :

programs for reducing scrams and safety system failures include effective  :

maintenance improvements.  :

The pie charts in Figures 4.3A and 4.38 show the distribution of the seven  :

cause codes for the subset of LERs in the PI program for 1987 and 1988. M1 is:

remaining around 28 percent, and M2 around 15 percent. These figures indicate:

that the fraction of maintenance errors in the subset of events for these two :

l time frames, has remained unchanged. (These fractions can be compared with  :

l those that apply to the larger set of all LERs in.F.igures 2C and ED, where M1 :

) is around 31 percent and M2 about 11 percent.)  :

2.6 Review of PI Data on Forced Outage Rate and Equipment Forced Outages  :

There are seven indicators in the NRC Performance Indicator Program for  :

operating nuclear plants: (1) automatic scrams while critical, (2) safety  :

system actuations, (3) significant events, (4) safety system failures,  : 4 (5) forced outage rate, (6) ecuipment forced outages per 1000jeritical hours, :

and (7) collective radiation exposures. Figure 5 shows the annual industry :

I 1 E

aD '

s -.

r, m

t o s s r a

D r

• A e q u

c S s N S S s 2 E i i

r G o r

E TA t L E a to r

t a

p o

r t

o S t a

M A

R G

O VAg g 8 8

2 D

a u

l R

P Pi A Vg 1

8 8

N AI 1

4. S T 4 EN RE UV GE I

F

$a 7 et 8r r

a u

O 1

r E 3 a C - e

. 7 Y N f 8 _.

A -

N _

E .

T _

N 2 I

A M fv g8 7 8

Ml3 1

7

( 8

- - ~ - - - - -

o o 0 0 0 0 0 . _

s "

0 8 2 0 8 s _.

2 1 i 1 1 _.

Cq) _.

1 , l I

N S S S E

G k

E Nb L

M A

R G  : : '

O R

P PI I

N T -

N E g V

E  :

1 E -

C N

A '

N E

T N

I A

M -

~-

7 0 0 e 0

5 4 2 5

N S S S S E

G E

L Ed

_ M R

R G

O A

l-P i

P N

I A

2 S

4 T E N R E U V G E I

F 2

E C

N A

N l"i E

T N

I b

A /

M k

_ 0 g 0 0 0

_ 8 , 6 5 4 5n o

S S E c N S S S S E

G E

L [M!!

M -

A R

G O

R P _

Pi N

I S

N E

V E

2 E _

C 5 N

A N

E T

N I

A M .

E L8 5 0 5 0 5 3 3 2 2 1 3SD

5 .

6 1

P S r T

N 1 e E .

h  %

V t t 4 E i n O 4 I

a 1 P M F 2 O  % .

7 t N . n AO 8 i a

3.T 87 2 I

4 U9 M

1 E

RB UI R GROF FT I

S I

D E  %

D 4  %

O 1 7

C  % .

2 6 E .

S 9

3 7 n p U

8 1  % g O

1 A t s

6 .

i s

C n u s 2 e .

o C

i n .

D i c

f o i p L b

r e m i u

m d q ...

N u

A E

1 1

P r  %

S 3 e

_ T 1 .

N h 5 t

E .

V t

n O 1 E i 2

I a

P M t F) i n

ORE S  % a T

6 M NR .

8O A U

7 3.

I Q 2 4 T E

UO R BWT U

I G

FI R T (8 8

S9 I

1 D R  %  %

EF O 5 . 7 D 8 O

% 1 p C 9 O E 7 S 0 n c U

6 6

1

% g i A t s .

2 i s L C n u s e o i .

C f

n i p D i

o u

b r

e m q m d E ...

N u

A

FIGURE 5 Annual Industry Averages, 1985 - 1988 Excluding Plants in Extended Shutdown Mode l l Cr d 2. Safety System Actuations e j

6. 53.

s 2r 2 '

lli 4 2.

t lis 1985 1986 1987 1988 t 1985 1986 1987 1988

3. Sigraficant Events 4. Sofety System Falures 2 C 3.

[e 6 _

z 2 og E2. 4 f gg .

347 2.3 2.21

< I- c.as o ss 2 2-0 0 1985 1986 1987 1988 1985 1986 1987 1988

's

. , FCURE 5 (COMID)

Annual Industry Averages, 1985 - 1988 Excluding Plants in Extended Shutdown Mode

5. Forced Outcge Rote (s) bYCrtYe$s No s 35 _

30 _ y 3_

25 ~

5 g 7 20 . 2 5 h.

' 15 . }

10 _ 1 o so 0 0 1985 19 8' 6 19 8' 7 1988 19' 8 5 1986 1987 1988 Yeor Year

( N) 8. Unit Avoimility Factor em _ g 80 ,.,

E k 600 . sn j 60_

3 9 400_ h 40_

200_ 20 .

0 0 1985 1986 1987 1988 1985 1986 1987 1988 i Year Year a.. . . . - - . . . . .. . . -m.......-.. ... . .. . . . . . . . . . _ _ . . .. . .

~.

.. I average value for each of the seven pIs and the unit availability factor, for :

the period covering 1985 to 1988. This figure is reproduced here from the PI :

Quarterly Report of February 1989.  :

Of the seven indicators those for forced outage rate and e  :

outages per 1000 critical hours (Figure 5, Blocks 5 and 6)quipment forced to :

are considered 1

be the ones most closely associated with overall plant maintenance. Forced  :

outage rate (F0P) is defined as the number of forced outage hours divided by :

the sum of forced outage hours and generator on-line hours. Equipment forced :

outages per 1000 critical hours (EF0) is the inverse of the mean time between :

forced outages caused by equipment failures. The mean time is equal to the :

number of hours the reactor is crdical in a period divided by the number of :

forced outages caused by equipment failures in that period. Both these  :

lodicators are used only for plants that are in commercial operation. To the :

extent these irdicators reflect maintenance activities, there is a  :

recognizable improvement in the 1988 industry performance in the area of  :

maintenance compared to the previous 3 years,  :

Figure 5, Block 8 shows unit availability factor for 1985 to 1988. Unit  :

availability is not a performance indicator. Data was obtained from  :

NUREG 0020 and shows an increasing trend consistent with declining forced  :

outage rates and equipment forced outages per 1000 critical hours.  :

2.7. Economic Assessment of the Cost of Maintenance Deficiencies This section presents rough estimates of the magnitude and trends in the cost of maintenance deficiencies to the commercial nuclear power industry. These estimates represent an attempt to bracket maintenance costs and their refinement will be the subject of further study as the factors influencing maintenance costs are beyond the scope of this paper. The results of these estimates are then compared with estimates in two other recent papers on the subject and the differences among the estimates delineated.

Two approaches were used to develop estimates of the cost and trends of maintenance deficiencies. One approach is based on the cost of replacement power during forced outages. Since the agency data bases on the root causes of forced outages are riot precise, a range in cost estimates is presented as a range of cost estimates for maintenance deficiencies. The upper end of this estimate is that the forced outages due to maintenance for all commercial ruclear plants have been between 50,000 and 60,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> per annum in the years 1981 through 1987. In 1987 this upper-bound estimate for these maintenance related forced outage was 54,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br />, but dropped significantly to a i projected 39,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> for 1988. The replacement power costs for these upper-bound estimates are 960 million dollars in 1987 and a projected 720 million dollars in 1988. The lower-bound estimates are considerably smaller being 120 million dollars in 1987 and a projected 80 million dollars in 1988.

Another approach examined forced outages associated with maintenance related trips. These outages were estimated to be running at the rate of 115 million dollars per year between mid-1987 and mid-1988. The trend in; replacement power l

costs for PWRs was flat in this time period while the similar costs for BWRs fluctuated widely.

I

'. . For comparison purposes, the draft Regulatory Analysis for the proposed maintenance rule estimates the cost savings associated with reduced downtime due to improvements in maintenance range from 159 to 239 million dollars per year.

2.7.1 The Trend in the Economic Costs of Maintenance Related Trips The cost of replacement pcwer for forced outages associated with automatic scrams which can be attributable to maintenance deficiencies were estimated from the automatic scrams tracked by the Performance Indicator program, the licensee event reports contained in SCSS, and outage time from the SINET data base. While it is recognized that forced outages associated with automatic scrams due to maintenance would only be a portion of the consequences of maintenance deficiencies, the agency data on scrams is complete and can be used to establish trends.

Automatic scrams which are attributable to maintenance deficiencies were estimated by the Nuclear Operations Analysis Center of Oak Ridge National Laboratories (ORNL) using SCSS to classify those automatic scrams attributable to maintenance deficiencies.* Forced outage times were then associated with individual maintenance-related scrams using the SINET data base. Finally, the outage times were converted to replacement power costs on a reactor by reactor basis using seasonal replacement power costs provided in NUREG/CR-4012.**

Using the methodology described above, the total cost of maintenance-related 2

scram forced outages *** was estimated to be $85.1 millicn dollars for the 9-month period beginning with the third quarter of 1987, which represents 5146.8 forced outage hours.**** (Theaveragereplacementpowercostforthese events is $404,000 per day per reactor.) Of this $85.1 million replacement power costs, $44.6 million (52.4% of the total) is attributed to BWRs and $41.5 l million (47.6%) to PWRs. Figure 6 presents the cost of these maintenance forced outages for PWRs and BWRs on a quarter by quarter basis for this time period. The cost of these maintenance forced outages was comparatively constant for PWRs, varying from $12.4 to $14.8 million per quarter. On the other hand, the maintenance outage cost for BWRs varied from $18.11 million in the third quarter of 1987 down to $6.8 million for the fourth quarter and up to

$19.7 million for the first quarter of 1988. Thus, on the basis of these quarterly replacement cost estimates for maintenance outages, one cannot identify a trend for BWRs and the trend for PWRs is level.

• Transmittal letter, G. T. May (ORNL) to S. Stern (AE0D), LERs on Maintenance-Related Problems, Nov. 7, 1988.

• • NUREG/CR-4012, Replacement Energy Costs for Nuclear Electricity Generating Units in the United States: 1987-1981, Argonne National Laboratories /NRC.

• • • The costs of these forced outages was taken to be the replacement power cost, only. NUREG/CR-4627 estimates restart costs from cold shutdown as less than $20,000 which is much less than the typical replacement power cost.

• • • • The PI data for the second quarter of 1988 was incomplete at the time this analysis was performed. .-

ll 11 s - -

t s _

o C 8 8

9 r 1 ee Q -

wcn I -

oa Pn e -

t t n nia .

eM .

s mt e

o 7

e e 8 r

u cadu 9 1

gl Q i

F ps 4 e ma .

2 Rr c _ 1 0

4 S e S R S

- C S g s: _

/

C ae l g '

G E ni a s R

> a ut r i r Ng OuO l l

a o

7 8

ts P o

r D 9 s d 1 oI C l' e f o Q r y e c s 3 el w d r n oe o i o Pi f t i F l l

i n

ea ss M

- - - - - p nl r C 5 0 5 e 0 5 0 5 0 cs 3 3 2 2 1 1 at l

n pe ev t

I E

~I .

. . 2.7.2 The Economic Costs of Maintenance Related Forced Outages The annual cost of replacement power for forced outages associated with maintenance deficiencies were estimated from forced outage times for operating plants stored in the SINET data base for 1981 through 1988.* SINET classifies forced outages as being due to " maintenance and test," and " equipment failure,"

as well as six other categories. All forced outages associated with the maintenance and test category are clearly attributable to maintenance. A brief aucit of forced outage events associated with the equipment failure category indicates the majority of these events are also associated with maintenance deficiencies. 11erefore, for the purposes of this economic estimate, the forced outages associated with the SINET maintenance and test category will be taken as the lower-bound of the cost of maintenance deficiencies. The forced outages associated with both the SINET " equipment failures" and " maintenance and test" categories will be taken as the upper-bound of the cost of maintenance deficiencies.

Figure 7 shows the bounded estimate for the forced outages associated with maintenance deficiencies from 1981 to 1988, excluding those plants in long term shutdown. The upper-bound estimate concludts that maintenance related forced outages varied between 50,000 and 60,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> per year between 1981 and 1987.

However, the upper-bound estimate shows a considerable drop in 1988, going from 54,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> in 1987 to a projected 39,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> in 1988.** Using the $443,000 per day replacement power cost utilized in the draft Regulatory Analysis for the proposed Maintenance Rule, the upper-bound estimates for the impact of maintenance translate into 960 million dollars in 1987 and a projected 720 million dollars in 1988. The lower-bound of maintenance related forced outages shows an irregular drop from 11,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> in 1981 to a projected 4,400 hours0.00463 days <br />0.111 hours <br />6.613757e-4 weeks <br />1.522e-4 months <br /> in 1988. The replacement power costs associated with these lower bound estimates are 120 million dollars in 1987 and a projected 60 million dollars in 1988.

For reference purposes, Figure 8 shows these same bounded estimates for maintenance related forced outages with c11 plants including those in long term shutdown. The differences between Figures 7 and 8 are due to the practice of some units to consider their long term shutdowns as forced outages. SINET reflects the licensee categorization of their outages. The upper-bound of the outage estimate in Figure 8 shows a sharp drop in forced outage hours between 1986 and 1988. Much of this trend should be properly ascribed to restart of plants under long term shutdown or reclassification of their outages by the licensees. Figure 7 should be regarded as the more accurate reflection of maintenance outage trends.

• 1988 estimate is based on data for the first nine morths of 1988.

• • The projected estimate for 1988 is a linear extrapolation of the data for the first nine months of 1988.

m_ _ _ _ _ __ _ _ _ _ _ _ _ . _ _ _ . . _ _ _ _ _ _ _ . _

e l

l 1

l l

'O  !

'N

$ i T E U h.

t:t0 5 l d I- l O

~$

(

s 1 \\\ s-

~

a l

o C5

.. Ah A _  ;

C~ l 3 '

Oo x =

W~ $

~ c-O N e l

y am 6 gC*

L'

}

{\  ;

g I

, =m ,

Ga = I x

- t x i O $ 5 r-L4 o

\ C C N l

'A ~  ; O F t i E

E h !o L ea E si

=

a  ::: .,

g .-

a w a so

=

o b N z a  ?.

C ,

s  :

- 8 u

\ S -

1. I l

i "' 5 e e e e l e a e ,

y a g \

i

.t

e l

1 I

i l A d l_

C a c) o I 2

=

ii c= s R'NNNN'N'NN'91:

i s,

l

=

,c) - , .

=

M

- O 3 $ $

N% .: $ \ $ >5 z

L J Q

,C Q

~

O \

r-J U F 3 t o .

h f. 0 l

L .t @

5 m a

'in: = 8 W c

• T Z $e k y I .c e , , , ,

e O C C C C ":f C C v N '

u

- 2.7.3 Other Economic Estimates Associated with Maintenance for comparison purposes, two other economic estimates associated with maintenance are examined: the NRC draf t Regulatory Analysis for the proposed maintenance rule

• and a recent article on plant maintenance by Theodore Marston of EPRI.**

The estimated economic benefit to industry and rate payers in the draft Regulatory Analysis for the proposed maintenance rule is based on an postulated improvement of five percentage point improvement in annual capacity factor at 20 to 30 plants. The contention is that these plants have substantial room for improvement in maintenance and are experiencing poor capacity factor largely due to maintenance weaknesses. This 5 percentage points in capacity factor translates into 360 to 540 plant-days of increased power generation. The avoided cost of purchase power would be 159 to 239 million dollars per year.

(The draf t regulatory analysis then competes the present value of this annual cost savings discounted at 10% per annum over 30 years and concludes that savings to be 1.5 to 2.3 billion dollars.)***

The Marston article identifies an average increase in operations and maintenance cost of 60 percent in the last 5 years. Maintenance is identified as being a relatively constant 25 percent of operations, maintenance and fuel costs. Maintenance costs at individual units are identified as varying consi-derably from 9 to 45 million dollars per year. Marston compares 1985-86 plant operating data with average data in the 1968-84 period and concludes that plant capacity factor performance has improved by about 1.8 percentage points due to maintenance related improvements in steam generator reliability and thermal efficiency. This 1.8 percentage point improvement in capacity factor would translates into about 720 plant days of increased power generation for 110 plants currently operating or about 320 million dollars per annum.**** Thus using Marston's data one can conclude the value of historic improvements in maintenance at about 320 million dollars per annum or an average 3 million dollars per unit.

• "Draf t Regulatory Analysis for the Proposed Maintenance Rule," Office of Research, NRC (to be published).

• • Theodore U. Marston, " Plant maintenance trends in the United States,"

Nuclear News, December 1988.

• • • This present value technique can be used to translate the annual savings calculated to a present worth.

• • • • A December 22, 1988 memorandum to J. L Blaha, OEDO, from R. A. Hartfield, ARM, " Chairman's Office Questions on Plant Performance Improvements" identified ,

an industry wide 1% increase in plant capacity factor as having a cost savings i of 240 million dollars. This ARM estimate is higher than an approximately 180 J million dollars for a 1% increase in capacity factor from the assumption in the I draf t Regulatory Analysis. The analysis in this paper will use the same assumptions as the draft regulatory analysis for consistency.'-

l l

i

___________________-__-_a

l' ..

l- .

1. , 2.7.4 Comparison of Economic Estimates Associated with Maintenance The various cost estimates presented in this section are quite different in their focus. The Marston data yields an estimate of 350 million dollars as a cost saving for present plants associated with past efforts to improve maintenance, but does not address potential savings associated with future actions. The draft Regulatory Analysis presents an effort to realistically estimate the impact of the proposed Maintenance Rule by examining those plants with degraded capacity factors due to maintenance and estimating the impact of the proposed rule on improving the capacity factors for those plant (159 to 239 million dollars per year).

The upper bound cost estimates developed from forced outages in SINET run at an annual rate approaching I billion dollars. This estimate represents the potential savings if all forced outages identified with maintenance deficien-cies and equipment failures were eliminated by improved maintenance. Thus, this estimate differs from that in that of the draft Regulatory Analysis as it nakes no allowance for the impact of the proposed rule and only identifies the consequences of eliminating all forced outages of certain classes. The lower bound cost estimates developed from forced outages in SINET run at an annual rate of about 100 million dollars that isconsiderably lower.than the upper bound estimates. The lower bound estimate also represents a potential improve-ment, but this estimate is limited to those forced outages identified as maintenance related in SINET. The rationale for these lower bound and upper bound estimates is that the data in NUREG 0020 and SINET may not be '

consistently reported and the AE00 analysts believe that many forced outages identified in SINET as " equipment failures" may indeed be due, in part, to maintenance deficiencies.

Lastly, an estimate of the replacement power costs associated with forced outages due to trips caused by maintenance deficiencies ran at about 119 million dollars per year from mid-1987 to mid-1988. This estimate is clearly on the low side as trips are only one of many potential causes of maintenance related forced outages.

In summary, the cost of maintenance deficiencies to the utilities and the ratepayers is running in the hundreds of millions of dollars per annun: or multiple millions of dollars per annum per unit, while the licensee expenditures on maintenance is estimated to range from 9 to 45 million per annum per unit.

3. FINDINGS AND CONCLUSIONS Based on the reviews of operational experience, the following findings and conclusions are reached:

1. Maintenance-related problems continue to be identified in many systems and components in several operating nuclear plants. The type of components and systems involved, such as motor-operated valves, solenoid valves, plant air systems, and service water systems, point out the pervasiveness of the problem and the potential for common cause failure of redundant safety equipment and systems.

'. . 2. Inadequate preventive maintenance practices account for a number of the maintenance-related problems identified in the event reports end generic communication documents.

3. Deficient maintenance procedures are also a major contributor to maintenance problems.

4. Human errors occurring during maintenance activities are also identified as part of the problem, and they point out the need for improved planning and personnel training.

5. The fraction of LERs attributable to maintenance deficiencies has remained:

relatively unchanged since 1985.  :

6. The trends indicated by the maintenance-related cause codes show that  :

plarit maintenance-related problems experienced by nuclear plants are  :

decreasing slowly over the last 4 years.  :

7. The distribution of causes for events that are included in the PI program :

when compared to all events in the same period, is similar to those for all :

events reported by LERs.

8. Trending of maintenance-related PI data show a decrease in the number of  :

maintenance related problems in the fourth quarter of 1987, and smaller  :

subsequent decline in 1988.  :

9. Trends seen in the PI data on forced outages and equipment forced outages :

also depict improving industry performance from 1987 to 1988.  : 1

10. Overall conclusion reached in reviewing these trends is that industry  :

efforts in the area of maintenance has improved in 1988 when compared to  :

previous years.  :

11. The cost of purchased power to replace generation lost by forced outages associated with automatic scrams due to maintenance deficiencies is estimated to be at an annual rate of over 100 million dollars in the period from Mid-1987 to Mid 1988. No discernible trend of improvement in these purchase power costs were evident during this period.

12. An upper-bound estimate of the cost of replacement power for all maintenance related forced outages is 960 million dollars in 1987 and a projected 720 million dollars in 1988.

1

l( l liI I '

F. Q E

R 'K

- K O 3 S, RO g'  : E Y , UC Y R 2 ,I R

S R U , , O R S 1 E E E F 2 ,ELS4 NLE L

L E

F L B & E 1 OLDi L S P - k S Y I cal l

C S I I SH I N E l SI SI I N N T I A l l

W L i

CLI I O l W

A L VY u O R I LNLP H O L U AU C R A AI UA H R P M DU M B F HMMPY1 N B' S

N I

N .

O' D

E 1

L 4 l

t A E U T L P T R E M E, I I U

U A N I C W E -

/ F E M R M M C I A E I A' E 1 ,

V I L M 1 b

Y W W I R N C D 0 F F l E O l R P S A A l

i t S C a t C 11 R R E I S E E S G I D A A VNG k E N EI OON i l

I L I SV E L 4 SCI I F t I . O C Al D I P O .l 1

1 S I R I I 0AHI PS SA NI E Ui I V 0RI ECSI R ELF S I N U R 1 ESl TORE l il M M 0 EN E I D I PGH DI D NL K S1 k S DARRW T I I E E' Ot U A N LD ENRDSKNU S SV Ib WHO WL E L Ui l OE ENl L I

U D GI U H I RCCEOHD R nAA l

Ak E HUT SU T AE NSHWH RCV MP P CUC RW FL0AU A YC R I I NN I 0 L1 AED BSSF UD A I EU NS 1 EI E' CLE , DO FE HI I EW F XN Cl Nl Gl BPf i

EPfD DiRu NGSG A D N E L t AY I A I A NS D E ANI - AUP GiWI L AK T OS ED I L I DFERRP USK I DE 4 MAI LA N t E C L AOLGDI OACL l R EM P I LI Y E I EE D L l EAI A - T L I DY I ASL 4 # ERLENMNS CRBB E UKLEB E I ERt VU kUOI L A S:

l N NC S L ST EU RI YuBRPS SHDR l t ASOY L N PC UST L F R RI EE I A SBSLR U I YO SI I I L .E E U L A E G' EG l l f I N E ACA B LRS t l

,l i EE V1 P S YEPA EMAFiNiN l l l sal #

LI LNY NS S LI W I M FI F I AMI I l N l AEBI E E RAN AYRD S ,F RAS ,

l I M VV AC I El 0S l UFOEDS I Dl R l I

t A E EEME F l' u1 AMUI NI EI HT S I A N L< KRl LN l I UE1 GEE I BSPMCE E L I CI uR R RDC I DNAI P I L MURG

. N U E k E A pal FSAE 4 SOSI L DRA -

1U k H0u0H L MNuFYNDNSRAWO RUEF i P C1 l 1 1 C I I F OSI I AADWSC CI O E l t

b R

A I E A L 1

N N -

S D R 0T UI R G E T l A

Y 1 NR MNE D N I R N' l

E R

A tnAE OLP MAP OLO N I NU E l

u I b AL H EL V S C d L CPO CPR A T NI E L l I i P NI SA NA I ESRu U

X EGh VNM EGM VNI OI RE I A F. , G M I UERi l NER J t I OVON A LI I LI l E 1 B I L I I S RLCI AR ARU AR1 NA L U UANP F VU0 VUI LE1 F OE D T VI OS D1 S D SLFMO CH N F I F DD iD i EGAFS A O MDMO GId ESENESUl l I I l

t N DR H I EO Nl 1 EUI T EDI SDAO EU S LI RN I PARDTAR l H RI t H L E AFO D RU I RUSl I F1 T UC E C FR I NRELI EELI E U A I N U l l OEGT I EPI NSPI NSR EN SA F E PNA BI UAE UAE LGRI SN l l EUI L A - FI H - FI l -

il t J

UMMEN I U R - T L E MWR SCR SCUI SAE1 U N E DHLA L ADUEN1 UENI RT STI I D E SSl DUT 1 EEI DAI l D AI PSENSSN C HMl I SS 1 I EUORWOORWMEROY UE E UAAUEN 1 1 SFMM1 SMM1 SI 1 PCSPI R CDFMRI 1 1 R - -

E U U B S S 8 M o8 U 1 3 3 2 6 0 2 3 2 N/ N 0 0 0 0 0 1 1 1 2 V - - - - - - - - -

e2 E 5 5 5 d 5 5 5 5 S g/ P 8 8 8 i M 8 8 8 0 B a1 Y 0 U U l N N N N N P1 T B B B I I I I I I

~

ll l ,!l l l '

A G e' R hL V l1 SRE DR NWNCA L C r Et iCI A u l i i l

R 3  !

AEUOU M t l P W F AMM D S SSC1 I S SK MRO l

l i

l H

U P I N H U L C 1 I

/

M H

S  !

E Y A I

M A S l S i G O G D l C G l

l Y t l D D / R E C / U S U E E S C R E S l' E

1 4

1 L D l i

l t 1 C l S S S I J Y- U L E

C S E V W N G D 4 l TAI Y U A E l l 1 t uRNE I f I

V. L A

O E lACl Ul OE f I D l l L T I I l RI A I Y ET l

aS AY V N 1 t DH I I Rl S1 HL MS t Vb 0 I E EW I SEl El kE 4

M l l

l 0 R L SO AF D l l

UI 1 E l I

)

U 1 E R I 1 OPL1 U 1 I E 0t UE I y G A

UE ADl P A l l

I R1 #

Cb t

1 b I l I I 1 I R LHRl l 0UGEl I

U I l A L K U C AHUOU 1 L N l' l AR A. I L Dl S A YI S I I F I SI i N 1 I MI l I

U i E D I E I t D l T Y C R ul Stl E L DG Ri l I 1 ADD EuSI E UD Sb GI UiG s E/ AT URE A L SER - f 4 E I l

i F i A R S P HREVR I I E F I I 1 t 1 l I E D ll l CF POG AS -

FAEUI A l

1 A D0i 1 1

1 I I

I ED DI uN UME 0SU 8

L L. 1 S11 A K E E l E i l ED Wt RE D E l. 1 8

1 F V W SI LSl B l DR G C f C HR R AA L I u R Nil CTI gal l . 1 I E E l A AD U . k ,

MB ARMARCEU I E I L I 1 I u E R I A E UR T I R 1 l SC l SSE C t tl l

lI PE SPDNFD E F Sl All E i

l Nl A H l

NUl l i

S l t il C UUF OWO EWPE I I A I l Af l

A R L D A I ER I R T I t

N I N Cl l

l luf al E LX N S HDPSS 1 l C I E E ) l uEt 4 l R I EOY UDD ES CE 1 I A L I l 0l iF l i l l' l i I LOE I SNSU I l 1 U1 1 lt H k1 l l ul l i l B ,T RLI I AOiL l Hb EH M SAE

~

l I E l E kI l CAWI F S OiN l DS b A i - D ASR AL U EL DPEI R 1 T DM ShGSE1 M t IUI M EME GA AOl RBHAi E E u APA I I l b R l I P I S K 1 Dt RRAI I CRt Sl R MRAi l

.ll u Pl SI I SE C UGGPCSLIl Eul S l T E I L S 1 U R MAUREOI A LI E MEE AiPkAY l Li Y I E E l' I I l' 1 SPD L SSDI BRCWUWCS Ul I CRR E

t b

A D H L R I D l

Nl l S U E D K l AI R M R PE A A T F 0 A2 I 1 UT E E li RAO1 ED A E

I E H

U E

t l

RA' I t L

C l WASR SH C b ME D 4

1 i dRE E 0 u I R 1 1 S E A S PGRN F1 I R - AD I LD lY Et E l OS CM P RY El A Ll S CDl G L US OT I E R I A A l GL RI N E SW U E H RAl. NI I N O SF E VED FOFE I M SA I SA Nll l RCI

• OT S N BH T OS I i l

i UNT AEE Al UC A L k l SAA' l C

SRHI SE ll i D EA SM E R r HI l L b I U l G I LEI Ul a CS I I N W t 1

S l l i

F LEt Y R RE L ENI OR L T lAl GEC LU AV A OH RH MS l DI I

i l l

u L l AI PI C RI L Gl 1

l E uAUl ALiE CA R E T M A l l L L MNl I E ll G I R LD f val K R A 1 D I I i I SE R ME l S A DUl C a I E

1 lASSS Rl t E E l UD L AT i Al HI 2

I I

iU t

I I

1 AI El1 DRE l E lCG UU

( R E

R NCUl I AE R E. OA LR I l 1

I Ul R

E :

I 8 M l s o0 l J

3 I 2 / l t 9 1 5 l N/ t 2  :

. 3 3 3 4 5 6 /

9 -

e2 L d 5 5 5 5 S 5 S 5 g/ l u l t 8 8 8 U 8 U 8 a1 Y t i

t t

1 l H 1 t H H N 1 l

P1 1 l I I I I I I I l l

A DGSNSO iM P

L UCUAAR CMSSL1 BR4PNC.A OU50A RT 311 RDC T VM Ai DS l

GN ci MS l

1 N

E tl u

P M

U S L C E E Y E

/ I L l J E. S M R P I # l t P D E E I R u I I T I U V S T R, I S I L C I R L I D Y A  ?

E S E U P -

S B ti l S 1 1 P 1 1

l l

i I

J D

E D l D

S T E l

0 I E A S 1 PDS D C L l O R l 1 I E I

E A A T l E C l 1

R C R SiG l A hI l

0 S G S E

P BN t t

M 0 A C i G D

N R O 1 K l EI l NE I E S i

ACVR l

l I ll OL T I EN EEXAO VD I B S RS VO VL TI B I R

I O E RN OI L t A AR T AO RT AETSF t 1 MP BI PI VHADO i 1 l

G R D T D D I Ei l G

UD N E Y A 0N R A HT S N E R F E A I C 1 O A P ST I 5 NT I I C Z ODD C P l R

I A E RN E A T I NDE I lAA L C UU T L l NSANF P DE N CM l

LE AL I UF E NR A S E M UA L OOOE N Bfl A NS 1 l SO Q E DT NR I N E EM E C ER 4 E I AE SC l 4 Nt I

l l l E L B

F OE DE AD t

0 I

CEI T T DP H 1 F

lUU .

I A I b O N S iEM T N EE A0 R NA T

I U R IELPN L I AD E

L SRL SE i l MR P OU U E t ANLE I Wl L il P I O NS P I F I AT NS M t l S S I E E AEAOR SR ul M l l Y U AD 1

1. LV E I ERSRI E SGU BA E. 0R O D A t 1

l HGD ESN Rl 3 M I RE R AN 1 A E AA ULUI L I I

B EI E RI 1 V R L K GSERA WD

.N U MT M G S T AAULCRR I

M AE 1 O R J

A A U EY EF X EELACUT  !

RE B

t i l I DB 1 O E SL P ACS JD l E l t

b E S S N T E E A A A GV R 1

EU NL E H R 1 Q I A H I U AE S I V P ' 1 F N 0DGN 1

S S S S I 0 O M 0ANO Y EN 0 D F 1 S I E ENI I I 1 O 1 l N E. T D I L DI I I U C B Dq SA ,I R ET I l A U 1

U NRAAH R C O Et DECEl l CA P. SS UOCI i l R LI N1 I CC I L D M R RT LF P L ,A NNE VO E O E P F CME A NUAS RS I C11 NM AED P Y T NLOMN EI R E I U SE1 I R

• SOOI MED S i 1 AP RRSRR E

l E -

' NI R1 OT N NM HS R E Y E 1 T I T I ACNA I A I I Y T N PESW T AHC I E LS SI lMS NPR OE A YI OI EAS ET F A l B

l LNCF1 MNLI S O EG YD JuO - U l

RE I A UEA SN l I R1 R

A E.

H EMYD0NI N i

4 l S 1 i G I E A E

D PUTO0EI ORI MEE AUEA

1. uI t l

D R

F1 RI i R l. M I

l l CVR I

OA oat I I EUE

_ L T RI L1 D WRSDM A R PT I ST 1 P l

A PsASATERA Z EE 1 S I' SS0EM Ml UONEPENF l l I T A SP E NI I ERl L l

_ 3 T I I QPI BOPI D H UO T I PNWI I I S. #

_ R E

_ B

_ 8 M o0 U 4 5 Y 4 7 0 6 B N/ N 7 7 7 8 8 9 9 V

_ 9 - - - - - - - -

e2 E 5 5 S 5 5 5 5 5

_ g/ P 8 8 U 8 0 8 8 1 a1 Y N N N H N N N #

t

_ PI 1 I I I I I I I I

A A NC L .

L NO O L I I R A AI ,

H P M S G S SZ1 A L

E 1

U H F E

N ,

S U S T E

P L N B M A A l O N L l

l 4

C R P

/ E M E I E L E H R B I

V I I M

S S G T R I V Y S X D N l P

l i

S N R E E R i

l R RE S P 0Y DD DD I E Ei l

L E 1 B S E H EE S 1 l HET l E E E l ll E T E E HA V C

l I S DD HLF l O T l

. l i T WF E EE I S AB l T tis SM l N U L 0 . l I T

1 V l E SR SA0T I RE 1 L l C

1. S l EM FEE UR1 UD Ol l 3 D E AE S S AMH ORV BC OE R t

M GFI t A DS FF l M A C . D SDK D Rt J I R I S F t E l EC ADGE L ECA OI END 0t K ;E ENCD CE EA I B l RE l AGL AO FXL I N A AV '

M ll B l GI NE NI L 1 I O l

AE i

l AAT AS I

I I AB

( EZ I E I N AR BR D l l U

I Zi B R CEAE MP lC l Gl l VSF D RRl al l . A I D HE ENEE A l l DN AOt ERR RO E B R

OD AT ik lAl EVT C l DASNA FV 1 I il ,

ORI Hl# ARi I l H

i I t ECEA l BR NE F E El bD T. D l H D S S OEDl W O V NT l I AC I

I A.

l I

t E LI I S RAMI I I t lHESEE t K i l NS OSI ADLA T UT Rl LR l DC AA E MH l h b 1 I E l uI BI S ACT AAD WDl DE l

R GW

.F l 1 AiGGWi l RP T U LCl N l H 0l RY l

l i

1 f i c

l 1

N EGSL B E SE , FT AI A A J

! b E PI I LG AN VN POI l

l 1 AM E  ! S G1 LWE AiET I E l EAERN W SC I G RI 1 L OVMCGI H At P HN 1 L CHI US 0ET Sl L RE0I L. E. AEAVRT T APSS EI R A 1 A H

EI l R1 F1 1

l DT T E S) OE VT l i .

1 l R l GO 0 L FD CET , ell L AEWE EE i l

l l LEEFLEiO .S E RCANi l b ALV G

)

LI R G I RT I l

Hl VT NHS N El HM VULEA L N l l 1

E FiWl I EE lAROl ENiAR RASGG AST I I l S Al K l

l i I M l i l O i i

Kli IK T T L EID EENNGS0I HG iI 1 B A E WT l I l uI W J Gt VI A M

t Y 1 W1 I NI ME t OS i L RT SEROE RI S - A RT LS H I RRl b L l l

D EL I S I HOE N WPi

.H u L J l

l EE LAV K OEE DSUGSGT WI T OAOHt R 0. JMll I RE I O R UD 1 AI 1 I E FA I

HVOl OHt HADAI I 4 t l ENFEi l E

F WT DBI GI AI CEFSOL HHRF PWA L I ADl l

E I b SFD A

l i N T E . HON D A T S l l T D AV R l E I S I M O LN 0 I GJ Gl T O CI I

.G0 AI - HO T I 0 D t 1 V I C CY I E1 NA UT ND 2 0S WE RC V1 GI RAI RR i EEM 1

L1 E' l I RA A I T T VAALN E Y E OD AE RFR' SRNWWE R V. T l D I HC VS i I U HE DOL ES I Cl lI K LTC I S 1 OPA T L Y VE AuR r G I Bt #O S A S AR iEBO l T N I EU DOI WE ED COWR B EI T LR E R T l HI J T CA E FR Sl I T LEiGI A l I N RHRFR M A } HS I I D/S Y LA WI D UUO SD .

LW 1 AA D - S AFL P H l S R l

t S E T US FE C 0 EI l SF NAIF T D T EDHFF S E F 1 l OL I uE A HONI I DEDOOS D OC I B WI RGH J ESE E il H D I l i NA 1 Pl G T E HENGif l AT GFB FSR I I EA I I NO l

SRI R l i

i t l l

l OS0 Ct D M1 S E i l

l sl CkE l

l I AI L U1 Y AB AE t O l l l i i AEI K CT SA St C RRHkT J l R l l l J l l l Ct b E SSI S A I I l bSSR A A AHkE0I I i l l' l l

O0E NEl l Y E O 4 I 1 1 I AIs Al i iAI T i l

1 V L 1 R I 'RI D S RC i

I R l E i B S 8 N 0 1 7 U 1

00 I I

J 5 7 0 0 0 O 1 / I 0 o. 0 1 1 1 l 9 - - - - - -

e2 E 6 .

. 6 6 6 6 6 g/ P 0 ti 0 0 8 8 0 Y N t l a1 N P1 I t

I I l

I I H H N I I I l

j' l

]Jl1

+

.~

. ,A ,M , , , ,

-? 4 EHDI M22I N .S 2 EELL,RYMD O I

?

T KRO 1 R ,R , t 2 N NONAGR0L Y ,3 R ,

3P I

O S AHRNLR1O t I E1N Y S,A APF0R F BN AD P P R L

U ,-I NOEHN Ek

, DEN Rt W 2 W T N 2MONBOUi S

T Y

E M B NEEHE , S BA ODAL KNHE D G

4I I O OI M ,

N K E Y MSUOCWCN N I LSRRRR Y A R L N SI OAA P CLAEBEER L

P U

T A

S A

M RED.iRREU ER VDMSEBPD l

R G

A I .i l

S 2 M S R U W N lVA1 NR il I EAI E l'

F E

I T

N L E

E N N R O I

O M T P U C N U E O C I

. C A O E

/ V R L M P P E . 1 I

C I E 1 S V V V L N R L Y P P R B O I U S R R S S C F M

. E S D V R S E E L G -

E E T l i O N H C C T R S I R E I E E ER O T I T YT Y O l i I F S O E BAR T T F A NS D WA S F NE M N D D E NE N U EGN 1 I V O F OL SNU AO L I O I B E UI O 0T GA T T O AR U ARB 0 NV C L D CA ED I U O MP I EE DE RB R R R S NBR AT I I 1 I OD E O U HA WU G S N F E NT R .

D I DS I L O N N O SI S E ES I I C I A DA OCE ERSH T L NE RAR C MTE S E N DE AH R P NTE l i E C I NR OC AOLDT T N A

. A DF CI S NNBE A I E AO RC E OG0 K N N SC E LOR T TRN1 A E I

- HN l l

E EB NFPA E T A -

1 A G E M I I l i X L N M 1 N VA T FO ARECO I EE P1 SOR MDC B E A 0 LT RS -

F ND . T M 1 l N A CL DT AAl A uI M NW I AG ENNHAT 0 N E BA E O T UN CI E NI 0 I SR M L E I DI NOT T I U E AEi l

.N B KS RI K APNNMC D 1 VL O AU RSA HTI ARR A RLI 1 D R EA EEE NEALEI N EAA P LC FRL ESMF1 C I CVF E

t T T C B N N M I A A A EE T 1 L L T R F AE O O SI OO MT -

O D O D V YF T OAL CYI CY I R S GM T RA RC RC S 0 ENE UEI RAA OD RAA L1 UI T APRNtD Y.

OD' E O DTS OEAR T NC T N G RE Y SY F T A CUI C U'C I A R FEES OOAL T MAil I

l AOR AOR 1 HL I I GT EBO EBO S OL I A F E I t R B R B - CI NKLE RSDRt E E O A EAAI UEHEi FRM* FRM WT DF DECL LVATR O U O' OUO TF I I LOE I L AA SR SR I US LR AAYM'.

NSF NSF KR QE OT FVT OE OE CD I V T HEM I F .

I RONI RGNO LL ET U FRLOI S TPNOT PNORT A SMAU OEI NR A I I A I I N YV El 0C l LBLEE DMTSDMT STI B RI 0ASEK AOMl l E AELUAELOEO DB UAEVEI NERPE l R1 URRT URGP Nl L1 D VAI CT I I '

l GSSRGSSRRT At i I HARLDRAl US l I EYEOEYEOAE T 0 AOHWAPPR0QA 5 I DSRCDSRC1 S SS FCI BVUS1tE F 1 2 R - - -

E U U B S S 8 lM 8 8 o0 i 0 0 2 3 6 7 5 N/ N 1 1 1 1 1 1 2 9 -

e2 E 6 6 6 6 6 6 6 g/ P 0 8 0 0 8 8 8 a1 Y N N N N N H N P1 T I I I I I I I i! l l  ! l lj lI j! jlI (l! } l !f{lll jl lll L

N GO ZI PDK P N0 A0R A N LN T ODRO K I A N L A I R I S SSR RBT A P I HOAiO l OU Alt E ll OA R S PA FCCI l C I PS LSP I RC G Q

E I

Y N T E M J E t

A F 0 A P C, M S I

U G L L L E E B E E E

/ S N M

L P

L P

N L 4

l N P E

T I I ,

u O I I T S S S C T S lL L R I R R L L Y U H S i C E E b U D S M M R P P S M F

A S

R C S o -

E E G G H

CTi DE I C

t E R 1 R I

E l

l N l l

0l 1

l SU SE I I I i1 I AS AP S t S T K W EO LO i

A E u I SA EK RP PR O W V t S F EEE ERC C X P CT 0S i

SVL L D C EA HE NM I - l iN A

l. F I AWB AI FD

'O L I B O liSY L I

Y E I A SE I

V fSB PST SR LR . CE 0S T O L l I EU EA CE F

AR U l LEL 1 Y E iMD RP RS H EP S I MP k WRE lI i l US 0T VSE D R SES DT L li DE 1 O E ,V EY E E OD Pt l 1 EO EC L A FE l

L1 RG NI 0 S A FCN CE D , EI I I E A C S .C EO O ES FG AF RK I

A. i AS LkS Rt K SE OAE FA 0A l

E l l l l PM PU URS K R S 1 E DE l atiG S R ,E NR A UEGAl CI l K R Ai A

f l A 1 WF SLUl ElJ L I L T l K S ESS OO T I QI LABI O UE E 1 l A AE PMY NAI T RSTF SC l

l i

CEV 1 ES L NV t

L D E FNS , E S i I NN l

A uI LL uT UO V HENPliES l

CS I A4 H RD l A JSD FO LDCT OMl I E CS l t Ul MV Y N H OEE I EI D AE EE O I I E LSA 0I SS I ET I WE0 C L I I 1 H A 1 L U T A L N I

I M E iDSD a EYI CDD E RAiARu ,SA 11 T EX VE E EEK MA I ECl 0T Y Ml E t l M

E R l

I I ST GCRA Rt E.W C R I P I 0NT EuR I

T D t I I UL OSHEEI l CE CAIB 0.V L l

L Ei

.N b

i t NEEL RT ADCVBD L ES t l AF 1 O ESL RI E PREANAOAI F U I O R A HFHS LI OEAH A E

I I

CAI I EDF SF1 MALNORRll F I PCI CGPI E A DC L S N Y H B G O A 1 R EM I t S SI O HE T T I E M ET I NT T C E T l M E RA A O S ED Gl 1 L

A ET AET T UI R RNY I E AO VEL T N S DD I OI S E1 K I I

D Y. LSOSA FY EA P B R AT Rt E AR I

I A VHI SO SY L OS CR O DE S

FDO HL DAO 0F t E A I Z I A C' O Ei K Rli RS ALL GT OT C C t PG H A B,N T OI R T I 1

I CR I

mal l NI T N NI I 'l EGAN u YN i

l l

i I C l W AE AR WH iOP K

I NRO t l E l S

DRI iSY L <

K I NO O F PI I C I EN AS NE T I T OUSET LN I NI T N T F O ME A1 R tARI OSRI I LI HOUSED SAP UR Y MF; AAi i OLCSWNN l O AI QEC I ESM EPE CT I E BI O FG ET ET NU T O I T ABRR AM N L A OQ FC PD I F , FRAOPHM 0I CNNECI EO LUT SOE T - I N 1 K I OT L T C EC I HR PDCW RO R RMI AN ASI VI D l l i L SUEAOSE l

l I EO E AT JOY i J

t ET SI T Ait l I

- DEL MPT RW PT C0I B 0R Al SAN 1 E Ekil LRAR E OA l SONEEI D Ml SME l

l Cil i At l

t l

l SF I OR I

i EOSPD i

MT RI i

LNARO1 I E EPCOAI AD U. N EU.UiE0l D

A l

l ESC 1 S I

RiU ol OLiRY rl A l AHRMERl UT l NAHY X UY 6 1 I Lf I l NDT F SI R FWAI DPI SS ,' I F I SiAS I

R -

E l l

B S

. 8 IM l

o8 A 5 O 2 4 6 8 0 1 l / t I

• 4 4 4 4 5 5 9 '- - - - - - - -

e2 E 6 o 6 6 6 6 6 6 6 g/ P l i i t l u 8 8 0 8 8 8 a1 Y H l l l N H N H N 1 8

P1 -

T I I I I I I I I 1

! l ' l

4*

.4- W- 1 W- -Q> m

.j -* .

Z aZJ W

N m- qg 3g 2 1

l

: m- 3 )

. m' 2NC e a a CWO C

= ,

• *W W W mMw M C ig r. g WWW . O Ok U 6U U

• wwm- m E MW gW W- M Ngmf 64 a

' ZMWWW CW&Zg

• WNUZ ,Z m

• 3'e ND = C ~

4 CEN

m. m64 e um am3EW E C 3" WmZ Z w Z >WZW2 *2E46m4mQC w

Z w oWW G og>

4 MJ2>4MWWU6ZOJ6 EJ E424 M COsCQ 4 U Z

ZmwZ wWm4 5

O 2C2 Caz J'

A W4gAUA*QCmNum3ECK DWCCCW *JECwg4 4 M

agM 3 UQUC M

=

MwW CAs

.i J

w W 2 = w 6 2 g

• W' C m -

k U A J w C E E < C d' C 4 2 W. =

U. W W W z Z

% J '> d W C M E - G. trJ . A -

U w W' w w Z' g w > Z > w E m y E W J w J 3 Q A 2

> 3 4 3 X 6 M M 3 m E E E M C U > a 6 Q Z QA W m

J s C ZO > 'a O w Ce J 2O N W Q CfJ W *Q

• 24 E wW (4 -Q U

w WWW Uw C MW w> m Wm -v C

'a M ww w U J m &m 6 g H Em3 >m gW ZM -45 QQ Z A E 6 Q CWW WM Z WW C wWC w a m MUM Wmm En EQ WM Z WQg w 3 E Zw4 QW M W WW J3 U A

• O C ZW .

f CJ wm M J6 EA ZWW JZCC-W ww U

UJ 4

JW

= w 20 M 4 24 mu W

W W

=

WWU

_gCt w

=

= 2" f-Ww w

• C .m 2 4 QC AD 4 MQ E w m =m>

42 JW 6CW6 WM ME- 2 w2W g

EA 54E CUA W>

2J42>MW m WM WC 4

M Cw>

W

- g=CZ.

.C A U U C4C 00U CQ WM w Www3&WW A> Z m m2 &E 3 m 6W 223 >CWQ w u Qw. 42 C C y MQQ ZH WC 462.Q4X ZC Qw We A ' WWZ M 4 QUW C CZ

-J WW- > *m(WCW mc WW WM wp CW Mg Z6J CCO O

2 QCC C

QW

.2 JJW JWJ --W wee 3WQ eQ. Ma Z ZC wu- dm W

wEC

- U A

E CCW m_a 4 4 A WmJGUCC U dw CZ 13 m JUO W' WCW WWm2 Z 36 EE Em E = C W

.g g 'mZM 6mCCmw e w6 & w 2CU -Q WC wd ZwJ 4- g' W *E WCJL3 Zm- mmw Z W J J-wg3 WWA

• Em4 WCJ CW m =

MZ WW' >>

WE Q W3*E E*LUJZC WQW WWmMW Uw 4Q3w G ZW

• C

-CWWJ

>Z=U J

WW JCZ CCCZM SC QMWUH eCQ WWQJ= - > gmc yZ gum WJ ZU>4CWmW s w E W -

H J W m$6MC A* g OC J3N

• UJWE<

W E.O NAw 4 CW3 3C W ACW mwmW E &ZW CH WZ ME2-0 . Neem E CZJ

• Z J-C gw2 2>3mWdm W

wguM JZg CZ>>> J Wm M.JWWCW W Cem43 EWCCW CW 6w Q

• CEQ C=.

• C E

w. 4 m EOW COWCCCC 346CZC4K

>4 EW UmMZG Z>WO*'

Um ZCyM H A wmW Em&EDUA 63CWwmeA M m46m w3m3 W

J

& k W E M CD C 4 -

C CE

> W > MJ W z C Z Q

• - 42 E C I w Jm3 L 4 m = OMCJ.

C W ZW >> C 2m2-

>= >> MC Z =

QC 2

C m w J > CJ Ww 22=6 3cm J JC Um OCH w .Z >> & w W- ZZ Um =W

> w m Z E Xw E2W 3J e EQC W O eE J

JM E mWJ JmA m

M 3

eJ 4

c 2

wc k g -

Ze waQ*Q w CM 43 ,4 4 &C 4 W2 ZW Qw gm

>3 Om CME =2 J w WZ mM e W EWW J e4 C 42 =WAC Z >> AA3 W E g ECW Q>WJL wW WJ CC 3 6 w WW. W J M 4 C A 6 CM Z N ac M4 JD ZQ W C Ugm mQC We w wm g Q W- u WJ E_sy W.

AM CZ CM w

C .g W 4 W MC MS WWw m A 03WM MWW J CZO J

W zw wW zww A JQ>U CU4 4M34 ,

AE 46 WJU C wWJ3 UCC CWQU M EZ W4 ACE g CW4> ZEM ME*W l N

• wm Em Cm* Q 66>m M yAA AWO3 W

• C E em 2%

g t1 n n 4 N W

O W

w 4

3 4

m N

4 N

3

)

> 3 I t i e i 1 1 I I WN W 4 4 4 4 4 4 4 4 4 Aw - = w a w w a w w w

UE U T LT t GO A .G GP G S D

- L t C , CO C Y L

A , N A L C, UI P SM2 MH MI G CI A DPI E

V S L C S A C T E V E N V E L D ,

H A E R T 0 V T E G P A T i B

M Y R A l l S U T E W E C E P M E ,

/ F O E I I 1 M A - C I S H E S R I A A E 1 O V i W V S R T R l l R G D Y Z O E 0 H D A R S P M S 0 R E R C f E O D I E SR S N L AC -

NC E ED H A L l .

EE I N C SN C H 0EE I A. N t

T N NAF E T I

l l A l l N DT 1 WH LPOD SO I ROEiG l T A HAE UA Y l I I E ADD T l l B O C , C MP RR N l

NS DE E T ,BYL EDG CDT N A1 I A W

DYA D - NI EHO S

M H EBC I DRE HK T N F0DE AL I l I HC NE L N WLOT l N AEl 0NL RRD LuT I I .

OL I EA l SE A I I Wi b l 0COC EN l iBA HP I B ACl I ER WS T o 1 C FPI S l l DWO T O FI E EFVU D R E N OOF ENl E N AR Vf U RC ST ENE SCDE R RI Cl VI M I' SRl 4 OEUC RA SE ENEE NP0 '

AlR R EL l l

RiCA . EE l l E , VAZB O 1 MS l iSI D OD VS lA Ol I Gl I AWD LLY I T' ERT S CE F E 1

Ll M I t

l 0UN CT L ALLT I 0D l ENENPD l T AI V .', - HO1 PAE EE VI A O A1 4 E ST ERAI 0 l I A Y I I I HT KBI ENN C L . Y LI -

L .G R T DUCA A l i RVA I SI MSI REFEt DE Y Sl l

ODALUXM EPS . HRND FRAE .FEGODF NR I I I I ERR EI L A I RUUA I OFLE VU A A ESl .CibIM E T GEN S H DT BGLDL GMS E F E SSAt l.

l l

l t T L S l EF N I CAOUAAEN R SC Al ENI AL OuAU WEEl ORAS LADRLOl DI DE l NN S I O HAE RE OiRt l f O R 0R I A E SECCROHE PE CI 1 NI LT C l T 5T E DR NFAL T N EE RC ENI E l SLS l

l Ai l T

FA D L ENHT EC NANEINE I

/ NEE GOSHFHI ALNF A . I I I L T ZA R I AI E0R BI CNl N1t A NCNT OCNR LN I L S R l i t G Ot ANR I E E1 DPt E :

UI M Rl MHEGDGI 3 4 E LT ECGNAT BA E il l EWVNI NMl D L CH I ELS RYCTCNl EBI G )e V RT EOA U

L M l SELOEI VI RI E ET LLT P T LEI l REAKC l

SVBNNTI T E LT Wt EAN L AE RT I LT E CR

.N C EROK El DTI UL Y LDESI EE WNDRST LI l A

I O R RiRNi EHEEOI l l ROl T I AHU l

EONOEEFI EiH l

I PSPUWSI SDFS DCASMMl F DDACl WOFRSC E

L B S 1 l N A Y E UYO 1 T SS T I T I S NI UR S SI 1 E AL AW E El l SL A E I CI =

2-C J 3 A Q

>=

2

.I C

A.

E O C =

W U N Z E

W *

>= .w I8) ()

> 1 C/J Z m > w W C 3 U ZLC.

M C

>= tra C

>= w W O CO@

E CCW 2W *= = 4 CJ Z2E

-& cO=

>= C UU%

CE EA e Cf3 3 2 U >=

CQ C4U gW

.>=- tra WC

-4 >= 2 2 J CCc QW JWW 2% WE C

W M

QE e W

C/) y

= .e .J >= v2C w3 J

m$

r*

CQ W e >=

WW L r= > f/3 J* C >= 4 W 2 W J e=

D~ E CMW

&C E.

W W

ZA2Q

~x ~

& 6- WZ=

• 2 O. m >= U M

== C M A

'>=WZC3 m W Q W

E-C

>= >=

CQ '

WW WP 24

• J 4W O3

• ts >=

2 >=

MW l.= 6 LfJ C y

W EtJ W W E 2W Wm-

>= lt)

C Ct) >

3 >= Lt3 CU W WCU  !

= - euW ,>

.-e

• 5 W

• IXl E

$ ED 2 PS

.a N Z CD

> e WN W CD A CD CP N.*

1e >= 2 LM >=' M

_ _ _ _ _ . _ _ _ - - _ _ _ _ _ . _ - _ _ - _ _ - - _ _ - - - _ - - _ - _ _ -