ML20126D891
| ML20126D891 | |
| Person / Time | |
|---|---|
| Issue date: | 07/31/1981 |
| From: | Jamarl Cummings NRC OFFICE OF INSPECTOR & AUDITOR (OIA) |
| To: | Bradford, Gilinsky, Palladino NRC COMMISSION (OCM) |
| Shared Package | |
| ML20126D877 | List: |
| References | |
| FOIA-85-55 NUDOCS 8506150236 | |
| Download: ML20126D891 (32) | |
Text
3 m
I fw fa aeg(o,,
t UNITED STATES
/
NUCLEAR REGULATORY COMMISSION
.y n
y
,I wasmucTow, o. c. 20ssb
- s., * "/
July 31, 1981 MEMORANDUM FOR:
Chaiman Palladino Commissioner Gilinsky Commissioner Bradford Commissioner Ahearne
[ _[
FROM:
James J. Cummings, Director Office of Inspector and Audito i
SUBJECT:
OIA REPORT ON NRR'S OPERATING MCTORS LICENSING ACTION PROGRAM Attached is our report on the Office of Nuclear Reactor Regulation's (NRR) program for processing operating reactors (OR) licensing actions.
Our review was initiated in response to increasing concerns over the number of OR licensing actions which had accumulated over the years, and the volume of new work being generated which, coupled with limited staff resources, would cause a dramatic increase in this backlog. Our objective was to identify potential improvements to more effectively manage the workload and enable more efficient use of resources throughout NRR.
-Our audit was perfomed during the period May through December 1980 and focused on NRR's management of the backlog of OR licensing actions, the prioritization of tasks within NRR, and the interface between NRR's licensing and technical divisions.
We also reviewed, to a lesser extent, NRR's management of tasks under private contract and the management infomation systems used throughout NRR for monitoring and controlling the work flow.
In conducting our review we contacted various representatives of NRR management and staff in three divisions involved in the review and processing of OR licensing actin'is, namely, the Division of Licensing (DOL), the Division of Engineering and the Division of Systems Integration.
We also attended a meeting between representatives of NRR and a private contractor, hired by NRR to review 500 pending OR actions, to discuss the perfomance of the contract. We did not directly contact any utilities, though we did review pertinent correspondence between NRR and selected utilities relating to various licensing actions.
Finally, we met with officials from the Office of Management and Program Analysis (NPA) to discuss sections of a draft of this report relating to management infomation systems.
Contact:
F. Herr, 01A T. Wiest, OIA 49-27051 5061g 6 850318 GRABER84-55 PDR
t s,
5
'i Th:t Commission 2
Following are the major findings of our review.
Summary of Findings The backlog of pending OR licensing actions rose to 3885 in February 1980, and gradually leveled off at approximately 3000 for the remainder of the year (2879 as of December 1980).
A high level D0L official told us that with additional plants coming on line, the backlog of OR actions could double in the next six years.
NRR management has established as a goal reducing this backlog-to not more than ten actions per plant, or a total of 720 actions for the 72 plants currently licensed, by the end of FY 84.
Although a number of efforts have been initiated to achieve this goal, our review dic. closed that these efforts have not been effectively implemented and that as a result, NRR has made little progress in tems of actually reducing the backlog.
OIA found the following to be the major factors impeding the reduction of the backlog:
1)
There is no formal, singular priority system that can be applied to all NRR work. The general guidance promulgated to the staff by the NRR Director in an April 2,1980, memorandum P act followed by the staff.
2)
The workload review that was initiated following NRR's April 1980 reorganization was never fully completed in tems of deleting completed or obsolete tasks, developing revised schedules for older tasks, and properly assigning tasks and providing work packages to reviewers.
3)
Although each DOL Operating Reactors Branch (ORB) prioritized the tasks assigned to it, no feedback was ever provided to the branch chiefs regarding how the tasks ranked with other ORB tasks or with other projects in NRR's workload.
Furthemore, the prioritization system was never perpetuated, but proved to be a one-time exercise to detemine what OR work was most important at the time of the NRR reorganization.
4)
Most technical division representatives were aware of D0L's effort to prioritize OR tasks, but told us they had not been provided with adequate priority guidance by DOL to detemine which backlog tasks to review first. Some technical division representatives stated that they were not even aware of any backlog tasks they were responsible for because DOL had not made fomal assignments or provided them with work packages.
As a result, the technical divisions are directing most of their efforts to areas other than OR tasks.
5)
There is no consistent perception of D0L's role in the NRR organization.
The confusion surrounds the issue of who is responsible for initiating backlog tasks as active reviews and whether priorities established by one division necessarily must be honored by others.
-. ~.
' :a 3
The Commission 6)
A somewhat adversary relationship between D0L and the technical divisions has hampered effective coordination and canmunica-tion among the divisions.
DOL feels it cannot manage tasks effectively without some degree of control over technical revi ewe rs. The technical divisions feel that DOL is often unrealistic and arbitrary in its schedule requests, so they make their own priority judgments.
There is little communication or feedback between staff members of the various divisions in terms of providing information required for reviews and providing feedback on the status of tasks.
7)
A better mechanism is needed for programming work and deter-mining the cost of work that is displaced to meet higher priority demands. The tendency to focus on immediate problems makes it difficult to anticipate future conflicts.
8)
NRR's contract with the Franklin Research Center (FRC) for the review of 500 OR actions was not adequately planned prior to its execution, and was not effectively managed in terms of dedicating adequate in-house staff and priority to support the efforts of the contractor.
As a result only 4 of 284 tasks assigned to FRC were completed by NRR as licensi.ng actions during the first year of the contract.
9)
Many of the NRR staff feel that the reports generated by the Regulatory Activities Manpower System (RAMS) are of little use in project monitoring and control because the reports contain infonnation that is incomplete and of ten inaccurate. OIA did not determine whether this condition has resulted from inadequate sof tware or from a failure of the staff to properly update the system with current and canplete information.
Conclusion Although the importance of backlog reduction has been well documented, little progress has been made in terms of actually reducing the number of pending OR actions.
It appears that no one has taken lead responsibility or developed a systematic plan for identifying and resolving the problems contributing to the backlog and accomplishing the stated goal of ten actions per plant.
Corrective efforts were initiated but have not been effectively implemented.
i We believe that if NRR canagement does not take steps to rectify the problems discussed in this report, the backlog will continue to climb as new regulatory requirements are developed and imposed and as new plants become operational.
In our opinion, a large backlog of unreviewed licensing actions has potentially serious effects on the NRC, the nuclear industry, and public health and safety.
i
~
c
~
The' Commission-4 Recommendations Based on the problems we identified during our review we recommend that:
~
- 1.
D0L take a more active role in monitoring the backlog and managing its reduction. This should include the designation of a specific. functional unit _ to have prime responsibility for reducing -the backlog, including reviewing all tasks to determine what action.is necessary, terminating obsolete tasks, finishing "near-completed" tasks, developing acceptance criteria, and
~
reassigning and developing revised schedules for the remaining t
~
tasks pending review.
~
2.
NRR develop a priority _ system encompassing all NRR work and which:
a) provides more definitive and realistic priority guidance in
' ~
terms of the demands being placed on the staff;
- b) assigns priorities to the work currently in progress and ranks items with the same priority and schedule according to I-
-relative importance; and c) provides a mechanism for prioritizing and integrating new work into the existing workload, including determining the i-
. cost and impact of ongoing work which is displaced when priority conflicts arise, and for periodically reviewing and revising e
established priorities.
e 3.-
NRR management develop a better definition of technical branch roles to eliminate the confusion among the 00L staff and_ the resulting delay in detennining where OR tasks should be assigned for review.
4.
NRR management assign a high priority to completion of proposed changes to 10 CFR 50.36 and 50.54 to reduce the number of incoming routine OR licensing actions and enable the. staff to devote more effort to matters directly related to safety.
In regard to the use of contractors to review licensing actions, NRR management must conmit' sufficient personnel to support the contract to t
- ensure timely and effective completion of the licensing actions. We 4
recomnend that NRR management:
i 5.
Develop general guidance to the staff on the handling of contractor work: a) stressing the importance of timely review of the contractor's' products and prompt completion of the lf corresponding licensing actions, and b) instructing the staff l;
to ascertain that adequate criteria and supporting documentation are available to the contractor when licensing actions are assigned.
l; i
s u
6 Tha Connissicn 5
6.
Dedicate the assigned project managers to the two FRC contracts and future contracts of this nature, on a full time basis to enable them to more effectively monitor contract perfomance and ensure adequate NRR support is being provided.
7.
Detemine the review approach to be used on multi-plant issues--
i.e., individual or generic reviews--before the work is assigned to the contractor. This will minimize waste of contractor effort and NRC resources resulting from changes in review approaches after work has been initiated.
Regarding NRR's computer and data needs, we recommend that:
8.
NRR and MPA jointly study the capabilities of RAMS to detemine how the system can be better utilized to produce reports that are more useful to the staff, and to detemine how NRR's reporting procedures can be improved to maintain the quality of the source data.
Agency Comments On July 23, 1981, the EDO provided comments on a draf t of this report (See Appendix I).
The EDO generally agreed that the report faiily represented the operating reactors licensing action situation during the period May to December 1980 when our review was conducted. He noted, however, that NRR was aware of the problems identified in the report and that corrective actions have been initiated, some of which began before the OIA audit was completed.
He stated specifically that:
(1) substantial improvements have been made in planning, assigning work, and tracking progress for those tasks assigned to the Franklin Research Center contract; and, (2) a Technical Assistance Task Force was established by NRR in May 1981 to review and assess NRR's technical assistance program and recomrend ways to improve it.
In responding to the specific recommendations included in the draf t report, the ED0 generally agreed with the thrust of all but one recommendation and stated that corrective action either had already been initiated or would be initiated shortly. The EDO stated that NRR does not believe that the NRR work prioritization system being developed should rank items with the same priority and schedule according to relative importance as we stated in our recommendation 2.b.
NRR does not believe that such ranking would yield benefits commensurate with the effort involved and that such decisions should be made on a case-by-case basis, recognizing the uniqueness of each item. They further stated that such ranking may not be possible since tasks within a given priority group may be of equal importance from a Commission perspective.
We recognize that such a priority system would be difficult to develop and implement. We continue to believe, however, that such a system can be beneficial to NRR and urge NRR not to discount the feasibility of such a system too early. NRR may want to consider including a study of the feasibility and possible approaches to developing such a system in its planned management study contract.
i
i V
4 6
The Commission In addition to the formal E00 conments on our report, we also received some verbal comments from the Office of Management and Program Analysis regarding Section VI of the report on Management Information Systems.
These comnents related largely to report balance and the feasibility of using the Licensing Actions Tracking System to provide on-line tracking of all NRR projects.
As a result of these conments, our draft report was revised to provide better balance between the capabilities and limitations of RAMS and how it can and should be used.
We believe the corrective actions outlined in the ED0's comments should resolve the problems identified in this report. We will, however, follow up on the conmitments made by the EDO in about six months to assure that they have been implemented and assess their effectiveness in resolving the problems in the operating reactors licensing action program.
Attachment:
As stated cc:
W. Dircks, EDO Office Directors Regional Directors 6
3 TABLE OF CONTENTS Page I
INTRODUCTION 1
Scope
1 Background
1 II ANALYSIS OF THE BACKLOG 3
Factors Contributing to a Large Backlog 5
Efforts Taken to Date 6
Conclusion 8
III PRIORITIZATION OF WORK WITHIN NRR 8
Conclusion 12 IV D0L/ TECHNICAL DIVISION INTERFACE 12 DOL Comments 14 Technical Division Canments 15 Conclusion 16 V
ACTIONS REVIEWED BY CONTRACT 17 Conclusion 21 VI MANAGEMENT INFORMATION SYSTEMS 21 Conclusion 22 VII RECOMMENDATIONS 22 VIII AGENCY COMMENTS 24 Appendix I ED0 Conments dtd 7/23/81 D
4 l.
I INTRODUCTION In May 1980 OIA initiated a review of the Office of Nuclear Reactor Regulation's (NRR) program and procedures for processing operating reactors (OR) licensing actions.
Our review was initiated in response to increasing concern over the number of licensing actions which had accumulated over the years, particularly since the accident at Three Mile Island (TMI).
Recognizing that as a result of TMI, management's priorities had changed and a multitude of new regulatory requirements were (and are still) being imposed on licensees, OIA was concerned that the volume of work being generated, coupled with limited staff resources, would cause a dramatic increase in the number of pending OR actions.
Our aim, therefore, was to review the OR licensing action program to identify potential improvements to more' effectively manage the workload and enable more efficient use of resources throughout NRR.
the In performing our review, 01A focused on three major areas:
management of the backlog of OR licensing actions, the prioritiza-tion of tasks within NRR, and the interface between NRR's licensing and technical divisions.
We also reviewed, to a lesser extent, NRR's management of tasks under private contract and the management information systems used throughout NRR for monitoring and controlling the work flow.
Scope DJr review was conducted during the period May through December 1980.
Our efforts were concentrated in three NRR divisions involved in the review and processing of OR licensing actions, namely, the Division of Licensing (00L), the Division of Engineering and the Division of Systems Integration.
We contacted various top manage-ment officials and staff members in all three divisions, including directors and a sample of deputy directors, assistant directors, branch chiefs, project managers and technical reviewers.
We also attended a meeting between representatives of NRR and the Franklin Research Center, a private contractor hired by NRC to review and evaluate 500 pending OR licensing actions, to discuss the performance of the contract. We did not directly contact any utilities, though we did review pertinent correspondence between NRR and selected utilities relating to various licensing actions.
Finally, we met with officials from the Office of tianagement and Program Analysis (MPA) to discuss sections of a draft of this report relating to management infonnation systems.
Background
Various events and activities require licensees to periodically submit information to NRC for review. Each of these submittals constitutes a licensing action, many of which are sanewhat routine in nature but nonetheless are required to keep the plant in opera-
- tion, i
a - --
2 There are four types of licensing actions:
- 1) amendments, which are formal actions resulting in changes to facility licenses, including changes to technical specifications; 2) the effort associated with public hearings which occur af ter an operating license has been issued; 3) letters sent to licensees requesting information that will resolve or aid in the resolution of technical issues; and 4) orders for modification of a license or exemptions to the regulations.
Licensing actions can apply to a single plant (plant-specific) or to a number of plants (multi-plant or generic).
Most licensing actions are initiated by NRC as a result of repula-tory requirements; operating experience; internal staff review of license applications or generic topics; or new information acquired through technological advancements, research or advanced analytical studies. However, licensees may also initiate actions to improve operational performance of their facilities.
10 CFR 50.90 requires that a licensee desiring to amend its nuclear power plant license must file an application with the Canmission.
Regardless of who initiates the action. NRC must review and process the submittal in order to assure adequate protection of public health and safety and the environment.
Historically as the number of licensed plants increased and as additional regulatory requirements were imposed, the number of licensing actions increased drenatically.
The prevailing philosophy was that operating plants had been extensively reviewed and found safe when they were licensed; therefore, staff efforts were devoted primarily to reviewing and licensing new plants.
As a result, those licensing actions not immediately needed in the interest of safety or to ensure continued operation of the plant were allowed to slip pending availability of staff to review them.
The number of OR licensing actions pending review rose to 3885 in February 1980 and gradually fell to approximately 3000 actions for the remainder of the year (2879 as of December 1980).
However, the Deputy Director, DOL, told us that with additional plants coming on line, the number of incoming OR actions could double in the next six years.
NRR management has recognized that clearing out this backlog of OR licensing actions is necessary. On several occasions management has stated that its goal is to reduce the backlog to not more than ten actions per plant, or a total of 720 actions for the 72 plants presently licensed, by the end of FY 84.
Eliminating the excess backlog (i.e., that over ten actions per plant) was stated as a major objective in NRR's FY 81 budget request.
This objective was also reflected in an April 2,1980, memorandum from the NRR Director to NRR Division and Program Office Directors which established operating reactors as the highest priority in the general order of work assigr.ments.
Again on July 22, 1980, the Director, NRR, affinned NRR's goal of reducing the backlog to a manageable level to enhance plant safety and minimize restrictions on plant operation.
Ir. spite of these statements, however, our review has shown that there has been little progress in effectively dealing with the backlog of OR actions.
t
'l 3
In March 1981 NRR revised its workload priorities in order to devote additional manpower to operating license reviews scheduled for completion over the next two years.
These plants have suffere&
delays as an indirect consequence of the TMI accident, which required NRC to focus its attention and resources on a reexamination of the regulatory structure and the development of new safety requirements for operating plants.
This shift in emphasis from operating reactors to licensing plants has deepened 01A's concerns regarding further accumulation of unreviewed OR actions in the backlog and their potential safety implications.
II ANALYSIS OF THE BACKLOG
~
In the course of our review it became apparent that the current backlog has yet to be adequately analyzed. While NRR officials are aware that the backlog consists of three kinds of actions--i.e.,
multi-plant actions applying either to all plants or to a group of plants; plant-specific actions, which may or may not be safety significant; and administrative actions, such as minor changes to or clarifications of technical specifications--they do not know a great deal about the substantive content of the actions in the backlog.
The only inventory of all outstanding licensing actions is a periodic listing prepared by MPA, based on input from NRR, in the
" Operating Reactors Licensing Actions Summary." However, the general feeling in NRR is that this list of actions is inaccurate and outdated.
Furthemore, this list identifies tasks by title and project control number but does not provide a definitive description of each task. As a result, the nature and scope of each task are not always readily discernable, nor is it apparent who among the technical staff is responsibile for reviewing it. MPA officials told us that the fomat of this list was designed based on NRR's stated needs and that any desired changes should be communicated to MPA.
Furthemore, they stated that the accuracy of the list is dependent on the accuracy of the data supplied by NRR.
We were told by several project managers and branch chiefs in DOL as well as technical division managers that many items in the backlog are nearly completed and require only minimal work to close them out, for example, by issuing the final letter to the licensee or by completing the fomal action to delete it from the backlog.
They also told us that many actions in the backlog are items that are no longer required, such as in cases where requirements have changed and the request for the licensing action is now obsolete.
They added that these actions continue to be carried in the backlog because it is felt that the time and effort required to close them out cannot be justified in view of higher priority work and limited staff resources, and that in many cases it would take as much effort to delete the action, due to the fomal procedures required l
(completing foms and obtaining concurrences), as it would take to f
l L
4 complete the review itself.
In addition, we.were told that NRC's statutory responsibility to protect public health and safety makes
-it difficult to delete actions without full reviews. NRR managers and staff both told us that as a result. these items are carried in the backlog from year to year rather than being deleted even though there may be no plans to complete the reviews.
We were told that other tasks in the backlog are inactive because the acceptance criteria needed to review them have not been developed.
On the other hand, one project manager told us that in effect there is no backlog because the important items are completed right away, and what is not important is deferred (i.e., becomes inactive) and will probably. never be reviewed.
A manager in one of the technical divisions shared this view, stating that DOL has shown little interest in completing older tasks in the backlog.
Some of the mid and lower level qanagers and staff we talked to--
those most directly involved in tne review of licensing actions--
pointed out that the backlog may be artificially inflated due to the method used for counting tasks.
Although many tasks can be completed with one broad review (multi-plant items), each task is counted separately in the backlog.
For example, a TMI Action Plan item may affect all 72 operating plants, but the resolution of the item may require that the same letter be sent to all licensees, rather than each plant requiring an individually-tailored response.
The difference between the two approaches in tems of manpower is significant.
For example, a generic letter to all licensees may roquire a single man-day to accomplish for all 72 plants whereas individual responses may require one man-day for each plant, or 72 man-days.
Although we realize there are few cases so simple and that mst multi-plant issues require some degree of review for each plant or class of plants, at the same time there are definite economies of scale to be realized on multi-plant issues when a broad review serves as the basis for completing several actions.
As a result of the confusion over the composition of the backlog, it is impossible to estimate the amount of effort required to reduce the backlog to the desired level, or whether the historical average of.12 man-years per action is realistic for projection pu rposes.
OIA could not detemine the impact of the backlog in terms of continued plant operation and public health and safety since these t
actions have not been fully reviewed. However, some NRR representa-tives believe that the impact is negligible since a large number of the pending actions are routine or administrative in nature and therefore, of low priority.
Furthemore, we were told that all new licensing actions which affect safety (such as unanticipated events) or continued plant operation (such as core reloads) are given highest priority by both DOL and technical division management and staff. However, a branch chief in one of the technical divisions stated that he does not believe it is necessarily true that the backlog items are not safety-significant to some degree.
He stated d
that a cursory screening by a project manager (when the schedule is set) is not sufficient for detemining a task's importance to
,e +
i y
safety, because project managers generally do not possess the degree of specialized technical expertise necessary to make such deteminations.
This branch chief maintains, therefore, that NRR cannot know the significance of the backlog with respect to safety until the tasks are given complete reviews by the appropriate technical experts.
Factors Contributing to a large Backlog We found several factors which have contributed to the large backlog of licensing actions.
The major factors include an increased number of plants and the resulting increase in workload from 4
licensing and maintaining those plants in operating condition; an increase in the ntsnber of regulatory requirements, many resulting from the THI Action Plan; and inadequate management attention devoted to getting the backlog reduced to a manageable level. This has all been compounded by prevailing manpower shortages and an inability to recruit qualified technical staff.
As previously noted, management historically has given higher priority to review work associated with licensing new plants rather ~
than to routine OR actions.
Although in April 1980 NRR management issued guidance which gives highest priority to operating reactors (see page 8 ), the staff told us that in practice this guidance is not realistic or consistent with the demands placed on them by management.
For example, the reviews for four recent near-tem operating licenses (NTOLs) in FY 80 required input from various technical disciplines. The high priority placed on these reviews required OR work that was ongoing or pending review by the technical staff to be displaced in order to complete the NT0L reviews on time. The staff feels that generally there is more adverse visibility from schedule slippages on licensing reviews than on OR actions.
Whether such " pressure" for licensing reviews is real or only perceived, the fact remains that it has adversely impacted the backlog of OR actions.
Since two of the four pending NT0Ls have now been issued as full power licenses, NRR management expects that more time will be available to review OR actions in the future.
I However, the several staff members told us that although less time will be required for NT0L reviews, they believe that other work continually arises that will be given a higher priority by management f
than the review of backlog items.
Especially in view of the recent push to complete OL reviews, it is doubtful that much progress will be made in the near future in tems of reducing the backlog.
Concurrent with the increase in the number of operating plants, the number of regulatory requirements has escalated over the years, bringing a corresponding increase in licensing actions.
Licensees are required to comply with regulatory changes ordered by the Commission, in keeping with 10 CFR 50.54 (h).
Among the more L
recent regulatory requirements are those imposed as a result of the TMI Action Plan, consisting of approximately 159 major requirements
.~
6 imposed on nearly all operating reactors. Many of these requirements, notably those highest in cost and complexity, will be carried in the backlog well beyond FY 82. The TMI Action Plan thus places an additional burden on an already overloaded technical staff.
We believe that a fundamental problem contributing to the backlog of OR actions is a lack of adequate attention by NRR management to resolving the problem.
As was pointed out before, even though OR actions were assigned the highest priority within NRR by the Director in April 1980, other work, i.e., casework, has in fact been given a higher priority.
In addition, although other efforts have been initiated, as discussed later in this section, we found that these efforts have had little impact on reducing the backlog.
As a final point, it should be noted that NRR management does not have total control over input to the licensing process, and hence over the backlog of OR actions.
The continual identification of new problems and safety concerns by the Commission, Congress and other interest groups has caused a continual shuffling of priorities.
As a result, many tasks must be deferred until staff become available to review them.
Furthermore, NRR management has little latitude in rejecting proposed assignments, given NRC's mission to protect public health and safety.
At the same time, however, this condition has adversely impacted management's efforts to control the backlog.
Efforts Taken to Date OIA recognizes that several efforts are underway in an attempt to reduce the backlog. That management has given recognition to the problem through the backlog reduction goals and the high priority given to OR licensing actions is in fact a step in the right direction.
However, the efforts taken to date have not proven effective in achieving management's stated goal.
In its FY 81 budget presentation NRR stated its goal of reducing the backlog to not more than ten actions per plant by the end of FY 84. This goal was reiterated in NRR's FY 82-83 budget submittal.
DIA was unable to find any documentation providing a rationale for this figure. Most of the branch chiefs and staff we talked to felt it was arbitrary and unrealistic, considering the volume of work being generated.
In fact, the ten per plant figure was established prior to the TMI accident and therefore gives no recognition of the number of licensing actions required by the TMI Action Plan.
However, some managers felt that since the volume of casework is not expected to increase much over the next few fiscal years, the technical staff will have more time available to devote to OR licensing actions and, hence, reducing the backlog, barring any unforeseen events like TMI.
Just following its April 1980 reorganization, NRR initiated an effort to review and prioritize the currently assigned work activities, '
and reassign all tasks based on the new functional units born by the reorganization. This was to be an iterative process which was
7 4
intended ultimately to enable a more orderly and effective flow of
- work throughout NRR, for both current work as well as anticipated futu re. work.
However, some technical division managers stated that DOL never followed through with the effort in tems of providing work packages and priority guidance once tasks were identified.
Other technical division managers stated they were generally aware of-the workload update effort but never received any task assign-ments from DOL as a result of the effort. OIA believes it would have been a worthwhile exercise if it had been fully executed and if the infonnation obtained had been communicated to the appropriate levels of management. As a result, many tasks continue to be " lost" in the backlog because they have not been properly reassigned or rescheduled based on their relative importance to other work activities.
(This effort is discussed more fully in Section III.)
Another effort intended to encourage NRR managers to devote adequate effort to backlog reduction was the establishment of specific goals as part of their SES perfonnance evaluation criteria.
Each manager would determine the number of actions he could realistically process, and this goal would appear in his SES contract.
Any variance would have to be justified, and exceeding the SES goal would be viewed favorably in the decision to award a bonus. While this approach is likely to bring positive results, a D0L manager stated that it will not solve the problem since it gives no recognition to the causes of the backlog.
In addition, he stated that the mere number of actions processed is meaningless since the scope and depth of tasks will vary significantly. As a result, managers can select actions which can be completed quickly to enable their SES quotas to be met.
A further backlog reduction effort undertaken by NRR was the awarding of a technical assistance contract in September 1979 to the Franklin Research Center (FRC) to review and evaluate approximately 500 pending licensing actions.
( A second FRC contract was awarded by NRC subsequent to our review for the review and evaluation of an additional 300 licensing actions.) The general NRR consensus is that FRC has done a credible job on the tasks they have completed.
However, the final safety evaluation reports and license amendments must be prepared by NRR, based on FRC's analysis. Delays at NRC on the first FRC contract have adversely affected the contractor's progress as well as completion of the licensing actions. The result is that little progress has been made in tems of further reducing the backlog.
(This contract.is discussed in Section V.)
Finally, NRR is considering changes to 10 CFR 50.36 and 50.54 which would enable licensees to make certain changes to their technical specifications without prior NRC approval. Present regulations require any changes to technical specifications, whether safety-related or administrative, to be reviewed by NRC and approved as a licensing action. According to some NRR representatives, a signifi-cant portion of the backlog consists of administrative actions.
Therefore, the proposed changes would reduce the number of incoming actions and avoid further growth of the backlog.
In addition, it would enable NRR to devote more time to matters directly related to safety.
'a 8
The proposed changes specifically would: (1) establish a standard for deciding which items derived from the safety analysis report must be incorporated into the technical specifications for a facility; (2) modify the definitions of categories of technical specifications to focus more directly on the aspects of reactor operation that are important to the protection of tM health and safety of the public; (3) define a new category of requirements that would be of lesser immediate importance to safety than technical specifications, thereby providing greater flexibility to both the NRC and licensees in processing proposed changes; and (4) establish appropriate conditions that must be met by licensees to make changes to the requirements in the new category without prior NRC approval.
An Advance Notice of Proposed Rulemaking inviting public comments was issued on June 30, 1980. The NRR representative coordinating the comments told us that licensees were strongly in favor of the proposed change. The target date for Commission review of the proposed rule is August 1981. OIA supports this effort as a positive step toward controlling input to the licensing process and ultimately reducing the backlog.
Conclusion Little progress has been made in tems of actually reducing the backlog.
Altrough the intent of many of the NRR actions--from assigning OR.sctions a high priority, to changing the regulations, to obtaining cor,tractual assistance on the review of some actions--
show a concern for reducing the backlog, these programs have not been effectively implemented.
Although the importance of backlog reduction has been well-documented, it appears that no one has taken lead responsibility or developed an overall plan for identifying and resolving the problems contributing to the backlog and accomplishing the stated gaal of ten outstanding actions per plant.
III PRIORITIZATION OF WORK WITHIN NRR In April 1980 NRR underwent a major reorganization which combined the technical support groups for operating reactors and casework (CP/0L reviews) to support all NRR work. The theory behind this pooling of resources was to ensure that NRR requirements were consistently applied to operating reactors and CP/0L reviews, and to promote more effective and efficient use of agency technical resources.
Because the technical review groups under the new organization had to support all NRR work, it was necessary to have some management guidance regarding the relative amount of effort that should be applied to each decision unit, i.e. operating reactors, CP/0L reviews, generic issues, etc.
Just prior to the reorganization, the Director, NRR, issued a memorandum on April 2,1980, to the NRR
s 9
Division Directors and Program Office Directors to define NRR workload priorities and manpower allocations for FY 80 and FY 81.*
The following order of priority was established to be used as a guideline for work assignments:
o Operating Reactors
.w o
Implementation of NT0L' requirements on Operating Reactors o
Unresolved Safety Issues o
Systematic Evaluation Program o
Implementation of NT0Ls for Operating License Casework for facilities with fuel load dates in FY 80 and FY 81 o
Priority I THI Tasks o
Operating License reviews for facilities with fuel load dates of FY 82 and beyond o
The six priority II and III tasks in the TMI Action Plan as designated in an April 1,1980, memorandum from the Director, NRR, to the Executive Dir~ector for Operations, "NRR Management Review of Draft 3 of TM1 Action Plan" o
Work in other Decision Units consistent with proposed repro-gramming shown in Enclosure 2 to April 2,1980, memorandum from the Director, NRR, to NRR Division Directors The new NRR organization als6 required a redistribution of assigned work activities from the previous functional managaisent units to the new units under the reorganization. On May 23, 1980, the Resources and Scheduling Branch, Program and Plorcir.g Analysis Staff, NRR, issued a memo to the NRR Division s ' Program Office-Directors concerning such redistribution or : M issignments.
All currently assigned items in the workloat i v1 as anticipated app...S ate branch or future items were to be reassigned to ths i
assistant director in the new organization.
The proposed reassign-ments were dictated by past and present functional assignments in the old and new organizations.
Each branch chief or assistant director was to verify that the tasks listed were correct; those in error were to be discussed and reassigned to another branch.
Tasks that had been completed or were no longer active were to be marked as completed or cancelled.
If possible a reviewer was to be assigned to each work item. At a minimum the branch chief responsible for the item was to be indicated.
- Subsequent to our review, the priority guidance was revised.
However, OIA does not believe that the revised guidance alters the findings and conclusions conveyed in this report.
.t of 10 Through an iterative process, i.e. reassigning tasks to other branches based.on feedback received at each iteration, the workload update ideally would enable each branch to know what work it was responsible for and put 00L in a better position to monitor work for progress.
Although the May 23, 1980, memorandum did not address all of NRR's workload, other assignments were made in separate transmittals.
Officials in DOL told us that in doing the workload update, project 4
managers were also requested to prioritize all current OR tasks, using a coding system from 1 to 3 (1 being the highest priority).
No fonnal criteria were provided for making these priority ranking detenninations, but project managers were allowed to use their knowledge and judgment in detennining what was important.
There was general agreement among the branch chiefs and project managers we talked to that the highest priority items were those which affected plant safety or the continued operation of the plant. Even though no fonnal guidance had been provided to the project managers, branch chiefs were quite confident and satisfied that the project managers' assessments and schedules were consistent with the importance of the tasks and reasonable in terms of time allowed to complete the reviews.
Project managers also felt that their experience and judgment enabled them to schedule effectively, though some felt that fonnal criteria were needed to ensure consistency with management's priorities.
Each OR task was to be reassigned and prioritized consistent with the May 23, 1980, memorandum.
Priority groupings were to be identified in a later document.
However, D0L branch chiefs told us they never received any feedback regarding how the priorities within their respective branches ranked with other ongoing work in 00L and in the technical divisions.
As a result the codes could only be used to detennine the relative priority of tasks within a single branch rather than for all OR tasks.
Furthennore, the prioritization l
system was never perpetuated, but proved to be merely a one-time l
exercise to detennine what OR work was most important at the time of the reorganization.
In fact, same management officials in the technical divisions told us they were unaware of the existance of the prioritization system, though they had heard that D0L planned l
to develop such a system. There was general agreement., however, l
that the priority codes would be helpful in conjunction with the l
task schedules in detennining where staff effort should be applied.
As a result of the failure to fully implement the prioritization system, the schedule set by the project manager remains as the sole indication of priority fnr new OR work.
For inactive tasks in the backlog--i.e., those not currently being worked on--for which the
- target completion dates initially set have passed, there is no indication of priority in relation to other backlog tasks or to incoming OR actions since revised schedules have not been developed.
/
11 Managers in one of the technical divisions knew which of the backlog tasks they were responsible for based on the f4RR workload update and their own research, but they added that they had not been provided by 00L with work packages or priority decisions necessary to know which tasks to work on first. They stated that from their experience, once needed OR work is identified, D0L fails to negotiate further in, tems of developing and communicating priorities and schedules and systematically managing tasks through to completion.
As a result of inadequate DOL guidance they have directed their efforts elsewhere rather than to the backlog tasks.
Assistant directors and branch chiefs in the other technical
~
division included in our review told us that while they were aware of the update effort and were told by D0L to expect a large number of OR task assigments, the assignments were never made.
They stated they generally knew of the backlog contents through the periodic listing in the " Operating Reactors Licensing Actions Summary," but that they had no knowledge of what the tasks consisted of substantively or whether the list was accurate. They added that with adequate infomation and guidance from DOL, they would willingly devote staff effort to completing any reviews assigned to them.
In the meantime they have channeled their efforts elsewhere.
There was general agreement in DOL as well as in the technical divisions that the schedules in themselves do not provide adequate infomation to detemine the relative importance of tasks.
Under the current " system" technical reviewers feel that schedules set by the project manager are largely arbitrary and unrealistic.
They believe that each project manager and D0L branch sets schedules, and hence priorities, relative to its own workload, with no regard for the demands placed on reviewers' time by other sources, including the Commission.
In theory, where conflicts arise regarding relative priorities and what is to be worked on at a given point, the conflict is supposed to be elevated to higher management.
Project managers told us that in reality technical reviewers merely ignore the schedules.
Technical review management, however, told us in the absence of definitive guidance from D0L, they are forced to rule themselves on relative priorities because they are serving all DOL needs (casework, OR actions, generic tasks, etc.), manpower availability is limited, and in most cases a short turnaround time is requested on all tasks.
In other words, technical reviewers and their management make their own judgments as to where manpower can be applied most effectively.
Given f4RR's limited technical resources it is unrealistic that all tasks with schedule conflicts could be elevated to higher management for resolution; consequently, only the highly visable tasks actually get elevated in practice.
Futhemore, a 00L branch chief told us that under the current system, managers beyond the branch chief level do not have enough infomation regarding tasks' relative need and impact on the staff to make decisions on priority. As a result, it is believed that the existing system for resolving priority conflicts is not effective because: (1) many tasks are never elevated; and (2) insufficient infomation is available to higher levels of management to make the decisions.
r 12 In the absense of a formal priority system that can be applied to all NRR work. OIA found that the approach currently used is some-what unstructured and undisciplined.
The previously mentioned April 2,1980, memorandum regarding priorities is not viewed by the staff as realistic, and is not followed in practice.
Furthermore, there is currently no mechanism by which workload priorities are periodically reviewed and revised to reflect new thinking.
Staff members also commented that because management's priorities are continually changing, it is of ten difficult to dedicate effort to tasks through to conpletion.
In their view, priorities change to meet the exigencies of the immediate situation.
In commenting on this, one project manager told us that due to the tendency to focus on immediate problems, there is little follow-up by manage-ment on overdue schedules except on highly visible tasks.
He added that he had not once been questioned by his management on tasks on which the schedules had slipped because management tends to focus on new problems that arise.
As tasks get older, their relative importance seems to diminish.
A technical review manager also told us that top management feels it has discharged its responsibility once a new task is assigned, and that it is left to the staff to see the task through to canpletion.
Conclusion We recognize that given NRC's mission, work priorities are very fluid due to the dynamic nature of the regulatory environment.
From operating experience, new research findings, and changes in political climate, relative priorities are constantly changing.
NRC management must remain flexible in order to respond accordingly.
However, we also feel based on our review, that management can improve the priority guidance that is currently available to the staff by developing a system that can be applied to all NRR work and which includes a mechanism by which priorities can be revised to reflect new thinking. More definitive guidance would enable D0L to set schedules that are more realistic and assist the staff in determining the relative importance of tasks where conflicts arise.
More broadly speaking, it would enable management to plan work items and utilize staff nore effectively.
IV D0L/ TECHNIC AL DIVISION INTERFACE From our discussions with NRR officials in the various divisions, we found that there was no consistent perception of the organizational role of DOL.
Some officials perceived all the NRR divisions as co-equals, each having " division stature", while others viewed the technical divisions purely as service organizations charged with providing technical support to D0L upon request.
In this latter view DOL functions organizationally as the lead division in NRR for routine operations, although it was acknowledged that occasionally situations arise in which the technical divisions function as co-equals, particularly where a division has responsibility to rectify
I 13 a safety problem.
In view of the leadership role they perceive DOL to be charged with, these latter officials believe it is 00L's responsibility to orchestrate the total NRR effort and to detennine how technical manpower should be utilized.
They added that particularly
-with respect to the backlog of OR licensing actions, 00L must take the lead in reducing it to a manageable level.
0IA did not attempt to detemine which perspective concerning the 00L/ technical review relationship is accurate. We found, however, that the confusion and disparity over DOL's perceived role in the
~
NRR organization nppears to have contributed to the problems associated with maintaining an orderly flow of work and with clearing out the backlog. The confusion surrounds the issue of who is responsible for initiating hacklog tasks as active reviews and whether priorities established by one division necessarily must be honored by others.
Our review revealed a somewhat adversary relationship between D0L and the techni: al divisions which has hampered effective coordination and communication among these divisions.
In the previous NRR organizational structure the licensing and operating reactors activities were perfomed by two separate divisions, the Division of Project Management (DPM) and the Division of Operating Reactors (00R), respectively.
D0R had its own internal technical staff to support the needs of operating reactors though technical assistance was sometimes obtained from the other divisions.
DPH casework (CP/0L reviews) was serviced by the technical divisions (Division of Systems Safety and Division of Site Safety and Environ-mental Analysis).
Although this structure eliminated the priority conflict between casework and OR work that prevailed in the organization under the Atomic Energy Commission (AEC), it was found to be operationally inefficient because it did not make the most effective use of NRC's technical resources, and because it fostered a lack of consistency in the review and application of new requirements on casework versus operating plants.
Overall it was thought that NRR could function more effectively by combining its technical resources, similar to the fomer AEC approach, to collectively service the needs of all licensing ma tters.
In April 1980 00L assumed the licensing responsibilities of the fomer DPM and DOR.
00L is organizationally divided into the following three functional groups, each managed by an Assistant Director: (1) Operating Reactors, (2) Licensing, and (3) Safety Assessment.
Each group is dependent to some extent on the technical divisions to carry out its mission and function effectively.
In the absence of any definitive guidelines regarding the desired allocation of technical l
manpower among the activities carried out by these three groups, or any workable and effective mechanism for resolving priority conflicts among the groups, technical review managers have made their own deteminations on the relative priorities of tasks sent to them for review. Our review disclosed that the allocation of effort in practice has not been consistent with the general priority guidance set forth in the NRR Director's April 2,1980, memorandum.
1.
^
14 D0L Comments Managers in the Operating Reactors Branches (ORB) under the Assistant Director for Licensing, with whom we talked stated it was difficult, if not impossible, for them to effectively manage their workload if they had 1) no control over the work input and 2) no control over technical reviewers. With regard to the first issue, not only do unanticipated reactor events (such as TM1 in the extreme case) interrupt ongoing work, but these mid level managers and staff claim-that NRR top management and the Commission continually make demands that disrupt and displace the efforts underway at any given time. They claim that the new demands continually placed on them by higher management preclude a commitment to a given task through to completion, and that priorities are too fluid as a result of the prevailing crisis-oriented atmosphere.
With regard to the second issue, an ORB representative stated it was his understanding that the new NRR organization would increase the role of project managers and give them more control over tasks.
Though this may have been one objective of the reorganization--that is, to promote stronger management of tasks--D0L representatives claim that it is impossible to manage tasks if they have no authority or control over technical reviewers.
They claim that technical reviewers do not honor the schedules requested by the project managers on OR tasks, and that, contrary to the NRR Director's guidance, casework receives highest priority by the technical divisions. They believe that one reason for this is that there is generally more adverse visability from a missed milestone on a casework schedule than on most OR actions. Although there is a postulated procedure for elevating schedule conflicts to higher management for resolution, as previously mentioned, in practice this procedure is seldom used beyond the branch chief level.
D0L ORB managers and staff also felt that the available tracking tools, in themselves, are not adequate for monitoring tasks (see Section VI) and that the infonnation provided by these tools must be supplemented by infonnal communication with the technical reviewers.
Project managers generally felt that the amount of feedback they get from technical reviewers assigned to their tasks is insufficient,
,^
and that of ten a technical reviewer will not even notify the project manager that a given task is expected to slip.
Project managers felt that technical reviewers generally do not focus on schedules or desired target dates.
Another concern expressed to us by D0L ORB managers and staff is that under the new NRR organization it is not always clear which technical branch has the specific expertise necessary to perform a particular review. The functional lines are not always well-defined, and we were told that in view of the overlapping branch roles the decision is sometimes dictated by which technical branch is capabie and willing to do the task within the desired timeframe, rather than a matter of clear-cut functional responsibility. Most tasks actually require input from more than one technical discipline,
^
15 and it is the project manager's discretion, with concurrence of his branch chief, to detemine the technical branches from which to request assistance.
Project managers told us that the infomal communication chain is the most effective means of detemining where tasks should be assigned, but added that better definition of technical branch roles is clearly needed.
Technical Division Comments Managers in the technical divisions with whom we met told us that 00L has not shown any interest in or taken the initiative to prioritize, schedule, develop acceptance criteria and initiate reviews on backlog tasks, and as a result the technical divisions are doing more casework and support work for other offices--Inspection and Enforcement and Nuclear Regulatory Research--than they otherwise would be doing. Because most technical review managers view their divisions as service organizations to D0L, they feel it is 00L's responsibility to perfom the basic work necessary to initiate these tasks as active reviews.
Representatives in one technical division stated they were unaware of any items in the backlog for which they were responsible. These representatives were aware of the listing in the " Operating Reactors Licensing Actions Summary" issued by MPA, but they stated that it is impossible to detemine the nature of a task by the title alone, and that they therefore had no way of detemining which tasks related to their areas of expertise.
They felt that at best, this list can only be used to project the workload since assignments must come from DOL.
They added that since they received no feedback in tems of task assignments from DOL after the workload update that occurred in May 1980, they assumed that tasks had been assigned elsewhere--for example, to the Operating Reactors Assessment Branch or to contractors.
Representatives of the other technical division in which we inter-l viewed stated that they were generally aware of the tasks in the backlog relating to their functional areas (based on the workload review just after the reorganization as well as their own research) but l
that they had not received (1) any guidance from 00L regarding how i
to rank the importance/ priority of these tasks, either by broad subject area or by individual task, or (2) the necessary infoma-r' tion to perfom the reviews.
They added that since many of the tasks have been in the backlog for so long, and with the frequent reshuffling of staff that has occurred, it was their belief that j
many of the project managers themselves were unaware of backlog items that pertained to their plants. These representatives further stated that they have not taken the initiative to pursue these backlog items in the absence of direction from DOL because they are short staffed and are already overloaded with other work.
Furthemore, they stated that DOL does not appear to be overly concerned about the backlog since they have not received any negative feedback from DOL concerning the relative amount of effort being applied to casework as opposed to OR work.
u
16 o
Representatives from both technical divisions in which we inter-viewed told us that they are receiving very little new OR work from DOL, but that on the assignments they have received they have tried to accommodate the project manager's needs, to the extent possible.
They added that sometimes the target completion dates requested by the project manager are not realistic, considering other demands placed on their time by the Commission and others, and that in some cases the target date requested is not consistent with the true need for the task.
From their view, project managers generally want a quick turn around on all their tasks and are not concerned with or knowledgeable of technical reviewers' commitments to other tasks or individuals.
Representatives of both the technical divisions shared the concern expressed to us by DOL representatives that management's priorities change too quickly to pemit dedication of staff to tasks to completion. They agreed that a better mechanism is needed for programming work assignments and for determining the cost of work that is displaced to meet higher priority commitments. They believe that more effective and realistic scheduling is needed, accomoanied by a stronger cammitment on the part of management to the established priorities to enable the staff to complete ongoing tasks before initiating new assignments.
~
Regarding monitoring and feedback to D0L, technical division representatives felt that the stronger, more effective project managers were better accommodated, in terms of acquiring technical time, than the more passive ones--that the attention given is largely a function of the individual. They generally characterized the licensing project managers (those assigned to casework) as being more demanding than the OR project managers, which further accounts for the fact that in the recent past more technical effort has been expended on casework than on OR work.
Conclusion i
Although NRR management has recognized the need to reduce the 4
backlog, recognition of the problem is not, in itself, enough.
It appears from our review that there has not been a systematic approach to reducing the backlog.
For the most part, each division seems to i
be relying on others to detemine which tasks should be done first, to collect the necessary infomation and to initiate the reviews as I
active tasks. 00L believes technical reviewers do not have time to devote to backlog items and in many cases are not even sure which functional unit should perfom the review. Technical review represen-tatives say they have time but need guidance and information, which must came fram D0L.
As a result of this confusion as to where the responsibility lies, much of the backlog has remained in an inactive status.
We believe that D0L, as the lead division for licensing matters, should take a more active role in coordinating and managing the backlog reduction effort. Though initial steps were taken at the
e 17 time of the NRR reorganization by means of the workload update, this effort was never carried through to completion in tenns of correcting errors in the backlog and properly assigning and developing new schedules for inactive tasks, and communicating this informa-tion to appropriate levels of management. As a result technical review efforts have been directed to other tasks where sufficient guidance and infonnation are available.
A certain amount of confusion and coordination problems may be typical in any "new" organization still in its transition stages.
However, the fact that the problems addressed in this report continue to prevail a year after the NRR reorganization suggests that they may be more deeply rooted.
We further believe that if NRR management does not recognize and take steps at this stage to rectify the problems we have discussed, the backlog will continue to climb as new regulatory requirements are developed and imposed and as new plants become operational.
In our opinion, a large backlog of unreviewed licensing actions has potentially serious effects on the NRC, the nuclear industry, and public health and safety.
V ACTIONS REVIEWED BY CONTRACT As previously noted in Section II of this report, NRC executed a technical assistance contract with the Franklin Research Center (FRC) of the Franklin Institute in September 1979 to assist in the review and evaluation of approximately 500 pending OR licensing actions.
(Subsequent to our review, NRC awarded a second FRC contract for the review and evaluation of an additional 300 licensing actions.) These actions resulted from various sources, including regulations requiring licensee submittals, operational problems and issues identified at specific plants, internal staff reviews of reactor safety issues, actions initiated by licensees to effect improvements in plant operations or efficiency, and from new information generated through research or technical advancement.
These 500 actions represented only a portion of the total backlog of OR licensing actions (over 2,500 actions when the contract was executed in September 1979 and currently near 3000 actions).
While FRC was contracted to provide technical support for the review and evaluation of these actions, the ultimate decisions regarding the acceptability of the proposed licensing actions were to be made by the NRR staff.
The contract was for one year on a cost-plus-award-fee basis, with two one-year options to be exercised at the Government's discretion.
The cost of the contract was estimated to be approximately $1.5 million for the first year.
FRC was to be assigned approximately 500 tasks (over three years).
The first year's assignments were grouped into nine generic task FRC was to detennine the extent to which the licensee's areas.
design criteria comply with criteria provided by the NRR staff, and to provide NRR with a written technical evaluation report (TER)
e'..
18 1
addressing the acceptability of the licensee's proposed design and technical specification modification. - All communications with the licensee, such as requests for additional infomation, were to come through the appropriate NRC project manager. NRR would review the TER at~ the draft stage and provide comments on FRC's evaluation and recommendation.
The final TER would then be used as the basis for the NRR staff's safety evaluation report (SER) in support of the -
proposed licensing action.
It.is generally felt within NRC that FRC's perfomance in the first year has been commendable in tems of its engineering and management accomplishments. Ironically, the major problem recognized by both parties to the contract during the first year of experience has
~
The NRC been not with the contractor but with delays at NRC.
project officer, responsible for overall management of the contract, told us he does not feel that NRR management has dedicated sufficient in-house resources to effectively monitor and control the contract.
He maintained that a commitment of at least one NRC technical manager (in addition to the overall project officer) for every five contractor employees is necessary to effectively manage and monitor the contract.
In practice the ratio has been closer to one to-3 ten. The effect of this has been an inability of NRC to support the volume of work perfomed by the contractor in a timely manner, delay in the completion of tasks and its resulting impact 'on the NRC and on licensees, and poor utilization of contractor personnel and hence,-of agency resources.
For example, during the first year of the contract (as of September 10,1980) a total of 284 tasks had been assigned to FRC.
Of these, 70 final TERs had been completed by FRC but NRC had only issued 4 of these as completed licensing actions according to FRC's monthly status report. We were told that six additional tasks did not require any fomal action by NRC and that a number of other tasks required only minimal work to complete, such as the final letter to the licensee or fomal action by the project manager to delete the task from the backlog.
However, until such action is completed, the tasks continue to be listed in the backlog.
NRC has held monthly progress meetings with FRC at which specific problems and policy questions have been discussed. DIA attended the meeting held on October 20, 1980, at NRC headquarters in Bethesda.- In that meeting an FRC representative commented on-the problems he had experienced due to NRC delay and the efforts he had taken to circumvent those problems.
For the most part his problems were caused by the NRR lead engineer's delay in reviewing requests for additional infomation fram the licensee (6 of 47 tasks) and in reviewing and providing feedback on draft TERs (28 of 47 tasks).
On one task, the contractor representative indicated he finally went directly to the licensee to obtain a diagram he had tried unsuccessfully to obtain through NRC (as required by the established procedure).
On 34 of the 47 tasks assigned to him, further FRC progress was dependent on NRC input.
L
e b
19 The NRC project officer told us that NRR's delay in reviewing FRC submittals is not the fault of any individual, but is an organiza-tional problem.
In the face of limited technical resources, other work has been given priority over FRC review work by NRR management.
The result is that NRC has spent a large amount of money on this contract, yet has made little progress in tenas of reducing the backlog of licensing actions.
In addition, the delay in reviewing these actions means that licensees continue to operate under outdated, inadequate technical specifications, according to the FRC representative and the NRC project officer.
The FRC representative also stated that NRC had not provided FRC with clear-cut acceptance criteria on which to base these reviews.
He stated that because the criteria were not well-defined he was uncertain whether the justifications he prepared would be judged acceptable by NRC.
Because many of the task evaluations assigned to him used the same justifications, it was possible that any error in judgment found by NRC would be carried through on all the tasks in that category, since he had received no feedback at that point in time. He added he found the lack of NRC feedback very frustrating and made it difficult for him to perfonn his reviews effectively.
The NRR lead engineer responsible for one of the generic task groups assigned to FRC told us that the lack of clear-cut acceptance criteria contributed to the delay problem at NRC in reviewing contractor work.
He believed NRR was not clear itself what it wanted FRC to do in the reviews. The lack of understanding and specificity when the tasks were initially assigned lef t NRR reviewers uncertain as to what to do with the information FRC provided based on its review. This lead engineer believed this problem is not limited to contractor work, but has long prevailed in NRR, and that the only solution is to make sure satisfactory acceptance criteria are developed before initiating a review. This is especially important when the review is to be performed by a contractor.
The NRR lead engineer responsible for another group of tasks told us that the initial infonnation NRR provided to FRC on some tasks was insufficient to initiate the reviews.
In most cases, the infonnation needed was available in NRC's files and was provided to FRC on request through written or telephone communications.
From our view, such correspondence led to delays which might have been avoided if the work packages had been throughly reviewed for completeness at the outset.
We were also told by an NRR branch chief that on some FRC infonnation requests, NRC did not respond in a timely manner and as a result FRC representatives eventually had to obtain the needed infonnation from NRC's Public Document Room before they could begin their reviews.
The NRC project officer's assessment of the FRC contract was thdt, for the most part, the contract has run smoothly and that the problems experienced were limited primarily to two out of nine generic areas.
In one of these areas, the problems were a result
L s.
,e 20 of unresponsiveness on the part of licensees in providing information
~
that had been requested by FRC through NRC.
In the other case, the problems were clearly the fault of NRC delay, as noted above. He told us the remaining seven areas presented no major problems.
However, OIA questions NRR's handling of two other task groups assigned to FRC. One of these task groups initially consisted of 47 tasks.
Shortly after being assigned to FRC, the entire set of tasks was put in a " hold" status by NRR management pending a decision to address them generically. That is, rather than reviewing each task individually, consideration was being given to developing a generic implementation letter to address the issue, requiring each affected licensee to respond, indicating and documenting their compliance.
In the course of their normal inspection activities, the Office of Inspection and Enforcement would audit a number of the plants to determine and verify that the criteria had in fact t
been met.
Because FRC had already initiated work on six of the tasks in this group, it was detemined that they should proceed only with those six, plus five other tasks in that group which Sandia Laboratories had previously analyzed and prepared reports on their adequacy. The remaining tasks would then be addressed generically.
Another set of tasks assigned to FRC for review consisted of 59 tasks. Five of these tasks were of a generic nature which had to be completed first to provide a basis for the other 54 reviews.
However, NRR management directed that 6 of the 54 tasks should be reviewed by FRC before the evaluation criteria were finalized.
Since FRC could not work on the other 48 tasks until the 5 generic reviews were completed, they were placed in a hold status and were still on hold at the completion of our rev'iew.
OIA questions the advisability of contracting out tasks for which the evaluation criteria have not been finalized. Obviously, any changes made in the final criteria may invalidate the conclusions reached on completed reviews and reouire tasks to be reanalyzed.
This is duplicative and an inefficient use of NRC's resources, and can be costly and frustrating to licensees as well.
Furthermore, we believe that FRC's efforts could have been more productively utilized by assigning tasks that were not dependent on completion of other reviews since available manpower could have been directed elsewhere.
In our view, NRR could have selected other tasks which I
were more appropriate and amenable to contractor review.
OIA believes the above to be an illustration of the poor planning devoted to this contract prior to its execution. Of the 284 tasks sent to FRC in the first year of the contract, 78, or nearly 30%,
were later withdrawn or deferred due either to a change in direction by NRR management or pending completion of criteria to review them.
With better planning NRR could have determined its preferred review approach and selected tasks for which evaluation criteria had already been finalized before assigning these tasks to the contractor.
This would have minimized inefficiency and enabled available FRC effort to be applied to other tasks.
N t
'21 In order for NRR to reach its goal of reducing the backlog.by 500 actions in three years through FRC' assistance, approximately 166 actions must be completed each year.
NRR was far from reaching this goal in the first year, with only 70 completed FRC reviews and 4 completed licensing actions.
Conclusion While we recognize that this was NRR's first experience using a
-private contractor. in the licensing process, we do not believe that the contractor was used to its fullest potential during the first year of experience.
Problems experienced in administering this contract, such as providing work packages to the contractor, failure to clearly define acceptance
.o criteria, not giving sufficient priority to the timely review of contractor products, and withdrawing tasks after assigning them to the contractor, indicate a lack of adequate planning by management in the decision to utilize a contractor.
In addition, management has failed to recognize that a contract cannot run by itself, and likewise 'cannot run effectively and efficiently unless sufficient in-house resources are dedicated to support the work of the con-tractor.This includes assigning a high enough priority to contractor work to enable its completion in a timely manner. Because NRR has detemined that completion of certain licensing actions is important enough to seek contractor assistance for their review, we believe management should also be willing to dedicate sufficient NRR effort to completing these tasks.
VI MANAGEMENT INFORMATION SYSTEMS The basis of NRR's management infomation system is the Regulatory Activities Manpower System (RAMS), which combines manpower tracking and project control to generate reports on manpower, scheduling, or many cabinations of both. NRR personnel receive a weekly RANS sheet and are required to charge their time to specific projects.
Each project is identified by a unique Technical Assignment Control System (TACS) number. The data is then used to generate various manpower reports, intended to be used to monitor and control individual tasks.
Several NRR officials told us that the various reports generated by RAMS ~ are of little value to them as task management and tracking tools because the infomation they contain is not always complete (e.g., all reviewer time charged to a task) and current (e.g.,
revised target dates).
Several of these NRR officials commented that they thought this was due to sof tware inadequacies which could easily be modified to better accommodate the needs of the staff.
Many project managers told us that they have developed their own systems to manually track licensing actions assigned to them because the RAMS reports are not responsive to their needs.
The vast
)
~-
22 majority of project managers we talked to agreed that in spite of the numerous reports generated by RANS, personal contact with the technical reviewer is the only effective means of monitoring their tasks, and that while they recognize that no report can substitute for personal contact, they believe that the RAMS reports could be modified to make them more useful as management tools.
One technical ' division director told us that because there is no singular management infomation system capable of serving the data needs of all' divisions, his division was attempting to develop its own tracking system for projects within that division, but that the system would not interface with the other NRR divisions.
In contrast,-HPA officials told us that the " inadequacies" of RAMS perceived by the NRR staff are due to:
- 1) NRR management in general not understanding the capabilities of the system; 2) a failure to r
routinely adhere to system reporting procedures to keep the system up to date; 3) a lack of-unifomity in the data neeos of the various NRR divisions regarding report fomat and type of-data desired; and, 4) inadequate planning by NRR for long-range data / reporting requirements, resulting in a fragmented approach to information reporting. MPA officials stressed that MPA is a service organization and developed RAMS in response to the stated needs of NRR; it was designed to accommodate the needs of the office as a whole rather than the needs of indiyiduals.
Furthemore, they told us that the reports are updated based on input from NRR.
As a result, they believe many of the inaccuracies are a result of NRR's failure to devote adequate time to maintaining the quality of the source data.
Conclusion OIA did not attempt to determine the cause of the problems noted above since that is not within the scope of this review. Our purpose here is to emphasize that the management infomation tools currently available are not found to be' satisfactory by many of the NRR staff.
We recognize the difficulty in finding a system and reporting format that meet the needs' of all divisions and individuals and that these needs may change periodically. However, the individual and division-wide systems have only marginal utility to NRR as a
~
whole, and have created a fragmentation in NRR's infomation reporting process.
In view of the interdependence, to some extent, of all NRR divisions, we believe there is a' need for an office-wide system that will meet the basic data needs of NRR collectively, and a need for improved procedures to ensure that the system is kept up to date with complete and accurate infomation.
VII RECOMNENDATIONS In spite of NRR management's recognition of the need to reduce the backlog of OR licensing actions, little progress has been made to date.
A major reason'for this lack 'of progress is NRR management's f ailure to effectively implement the corrective programs it has developed. Based on the problems we identified during our review we recommend that:
P
~, -, - -
23 1.
00L take a more active role in monitoring the backlog and managing its reduction.
This should include the designa-tion of a specific functional unit to have prime responsi-bility for reducing the backlog, including reviewing all tasks to detennine what action is necessary, teminating obsolete tasks, finishing "near-completed" tasks, developing acceptance criteria, and reassigning and developing revised schedules for the remaining tasks pending review.
2.
NRR develop a priority system encompassing all NRR work and which:
a) provides more definitive and realistic priority guidance in tenns of the demands being placed on the staff; 1
b) assigns priorities to the work currently in progress and ranks items with the same priority and schedule according to relative importance; and c)-provides a mechanism for prioritizing and integrating new work into the existing workload, including determining the cost and impact of ongoing work which is displaced when priority conflicts arise, and for periodically reviewing and revising established priorities.
3.
NRR management develop a better definition of technical branch roles to eliminate the confusion among the 00L staff and the resulting delay in detennining where OR
. tasks should be assigned for review.
4.
NRR management assign a high priority to completion of proposed changes to 10 CFR 50.36 and 50.54 to reduce the number of incoming routine OR licensing actions and enable the staff to devote more effort to matters directly related to safety.
In regard to the use of contractors to review licensing actions, NRR management must commit sufficient personnel to support the contract to ensure timely and effective coupletion of the licensing actions. We recommend that NRR management:
5.
Develop general guidance to the staff on the handling of contractor work:
a) stressing the importance of timely review of the contractor's products and prompt completion of the corresponding licensing actions, and b) instructing the staff to ascertain that adequate criteria and supporting documentation are available to the contractor when licensing actions are assigned.
~.
~
6.
Dedicate the assigned project managers to the two FRC contracts and future contracts of this nature, on a full
~
time basis to enable them to more effectively monitor contract perfonnance and ensure adequate NRR support is being provided.
7.
Detennine the review approach to be used on multi-plant issues--i.e. individual or generic reviews--before the -
work is assigned to the contractor. This will minimize waste of contractor effort and NRC resources resulting from changes in review approaches af ter work has been initiated.
Regarding NRR's computer and data needs, we recommend that:
8.
NRR and MPA jointly study the capabilities of RAMS to detennine how the system can be better utilized to produce I
reports that are more useful to the staff, and to detennine how NRR's rcporting procedures can be improved to maintain the quality of the source data.
VIII AGENCY COMMENTS On July 23, 1981, the EDO provided comments on a draft of this report (See Appendix I).
The ED0 generally cgreed that the report fairly represented the operating reactor licensing action situation during the period May to December 1980 when our review was conducted.
He noted, however, that NRR was aware of the problems identified in the report and that corrective actions have been initiated, some of which began before the OIA audit was completed. He stated specifically that:
(1) substantial improvements have been made in planning, assigning work, and tracking progress for those tasks assigned to the Franklin Research Center contract; and, (2) a Technical Assistance Task Force was established by NRR in May 1981 to review and assess NRR's technical assistance program and recommend ways to inprove i t.
In responding to the specific recommendations included in the draft report, the EDO generally agreed with the thrust of all but one l
recommendation and stated that corrective action either had already been initiated or would be initiated shortly. The E00 stated that NRR does not believe that the NRR work prioritization system being developed should rank items with the same priority and schedule i
according to relative importance as we stated in our recommendation 2.b.
NRR does not believe that'such ranking would yield benefits commensurate with the effort involved and that such decisions should be made on a case-by-case basis, recognizing the uniqueness of each iten.
They further stated that such ranking may not be possible since tasks within a given priority group may be of equal inportance from a Conmission perspective.
We recognize that such a priority system would be difficult to develop and implement. We continue to believe, however, that such a system can be beneficial to NRR and urge NRR not to discount the l
feasibility of such a system too early.
NRR may want to consider l
including a study of the feasibility and possible approaches to developing such a system in its planned management study contract.
25 In addition to the formal ED0 comments on our report, we also received some verbal comments from the Office of Management and Program Analysis regarding Section VI of the report on Management Information Systems. These comments related largely to report balance and the feasibility of using the Licensing Actions Tracking System to provide on-line tracking of all NRR projects.
As a result of these comments, our draf t report was revised to provide better balance between the capabilities and limitations of RAMS and how it can and should be used.
We believe the corrective actions outlined in the ED0's conments should resolve the problems identified in this report.
We will, however, follow up on the conmitments made by the E00 in about six months to assure that they have been implemented and assess their effectiveness in resolving the problems in the operating reactors licensing action program.
I i
4 4