Report of the PRESIDENTIAL COMMISSION on the Space Shuttle Challenger Accident

 

Volume 2: Appendix G - Human Factor Analysis

 

 

[G1] The Commission staff investigators reviewed the work schedules of NASA and contractor personnel involved in the launch processing of the Challenger at Kennedy and of the Marshall managers involved in the January 27th teleconference discussion of low temperature effects on the Solid Rocket Booster joint. The results of the review are presented herein. Although major accident investigations now include human factor analyses, the Commission avoided drawing specific conclusions regarding the effects of work schedules on work performance or management judgment. However, with the concurrence of NASA officials the Commission agreed that the results of the review should be included as an appendix to the Commission report. An evaluation by NASA of the consequences of work schedules should be conducted as part of its effort to reform its launch and operational procedures.

 

Workload Demands at Kennedy Space Center

The ground crews at Kennedy have final responsibility for physically preparing the Shuttle for launch. To meet the mission launch time requirements, the entire preparation sequence follows a strict schedule that imposes an around-the-clock work requirement on several large groups of workers including launch control center personnel.

While shiftwork is commonplace in many industrial settings, few can equal a Shuttle launch's potential for inducing pressure to work beyond reasonable overtime limits. Such workload pressure increases when launch rates are accelerated or when unanticipated delays are encountered in the launch schedule, especially within the last few days before lift off. Research has shown that when overtime becomes excessive, worker efficiency decreases and the potential for human error rises.1

 

Shiftwork, Overtime and Safety

The importance of shiftwork and overtime as safety issues is underscored by the variety of accidents and incidents which have been reported at Kennedy in recent years. The Lockheed Space Operations Company Incident/Error Review Board reviewed a total of 117 and 147 incidents (mishaps with property damage less than $25,000) during calendar years 1984 and 1985, respectively. Of these, 51.3 and 56.5 percent were attributable to some type of human error, including procedural deviations, miscommunications and safety violations.2 There were also 12 reportable mishaps with property damage greater than $25,000 in 1984 and 24 in 1985.3 The incident statistics, while revealing, underestimate the extent of the problem because some incidents never get reported. 4

One potentially catastrophic human error occurred 4 minutes, 55 seconds before the scheduled launch of mission 61-C on January 6, 1986. According to a Lockheed Space Operations Company Incident Report, 18,000 pounds of liquid oxygen were inadvertently drained from the Shuttle External Tank due to operator error. Fortunately, the liquid oxygen flow dropped the main engine inlet temperature below the acceptable limit causing a launch hold, but only 31 seconds before lift off. As the report states, "Had the mission not been scrubbed, the ability of the Orbiter to reach defined orbit may have been significantly impacted." 5

It wasn't until after the hold was called because of the temperature drop that anyone realized that liquid oxygen had been dumped. As Arnold Aldrich, Manager, National Space Transportation System, later commented, this loss of liquid oxygen "could have led to serious safety of flight consequences had the team elected to waive the SSME [Space Shuttle Main Engine] temperature redlines and proceed with the launch." 6 The investigation revealed that console operators in the launch control center at Kennedy had misinterpreted system error messages resulting from a failed microswitch on a replenishment valve. Consequently, instead of manually overriding an automatic sequencer and closing the next valve in the sequence, they pressed "Continue" and caused the vent and drain valves to open prematurely. The report cites operator fatigue as one of the major factors contributing to this incident. The operators had been on duty at the console for eleven hours during the third day of working 12-hour night (8:00 p.m. to 8:00 a.m.) shifts.

 

Excessive Shiftwork Pressures

There are over 5,000 contract workers engaged in Shuttle processing at Kennedy. These individuals are responsible for assembly, fueling, launch pad preparation and other pre-flight requirements. Mission 51-L was one of several missions they were processing during the three months preceding the accident on January 28, 1986. In January five aborts and two launches, of which three fell on Saturday, Sunday or Monday, resulted in an unusually high amount of overtime for these employees.7

During November, December and January, the average monthly overtime percentages for the various directorates at Kennedy rose significantly. Some were as high as 27.7 percent, and many were in the 20 to 26 percent range (48- to 50-hour weeks). During interviews both Robert Sieck, Director, Space Shuttle Operations, and David Owen, deputy program manager for Lockheed, have described 20 percent overtime as equivalent to a 48-hour or six-day week.8 Such descriptions can be misleading because they are based on averages that include all workers including those with no overtime hours. Thus, while they provide a rough estimate of overtime trends, they do not address the distribution of overtime within the group. This is important because it is individual performance, not group performance, that affects the safety of the operation.

Consequently, Commission investigators analyzed 41,105 individual weekly timecard records for all 429 Morton Thiokol employees from October 6, 1985 to February 2, 1986, and for 2,451 Lockheed employees from October 26, 1985 to January 31, 1986. The Lockheed sample included all workers in the directorates that had the highest overtime percentages during the three-month period under examination (Process Planning and Control, Orbiter Processing Facility Operations, Launch Control Center Operations, Pad Operations, Facility Operations and Maintenance, and Reliability and Quality Assurance). The analysis of both sets of records was designed to permit comparison of the actual overtime hours with the guidelines of management.

 

Management Controls

Both Morton Thiokol and Lockheed Space Operations Company have management policies regulating employee overtime.9 Each requires prior approval by the director for anyone working more than 20 hours, or 50 percent, overtime in one week; however, they differ in their regulation of successive workdays. Thiokol requires advance vice presidential approval for anyone working more than 13 consecutive days. Lockheed requires such approval for working more than six days in a week but has no formal policy governing more than six successive workdays if they [G2] occur across a two-week span. Instead, some directorates have informal policies or guidelines that are used to plan shiftwork scheduling. Neither company has any policy that regulates the number of successive overtime days separate from that governing successive workdays.

The individual weekly labor records revealed that the 20-hour limitation was exceeded 480 times by Morton Thiokol employees during the 18 weeks examined. More important is the distribution of this high overtime throughout the workforce. Only 169 of the 429 workers shared these weeks. Thus, although the overall average number of high overtime weeks was only 1.12 per employee, it was 2.84 weeks for those actually affected and ranged up to seven weeks for some individuals. The amount of weekly overtime often substantially exceeded 20 hours. In fact, there were 35 instances of employees working 40 or more hours overtime in addition to their usual 40 standard work hours. 10

The equivalent number of high overtime weeks for Lockheed employees was 2,512 during the 14 weeks examined. Again, only 1,013 of the 2,451 total employees were exposed to such high overtime, resulting in an overall average of 1.02 weeks per employee but an average of 2.48 weeks for those actually affected. The highest incidence occurred during the week of November 9 to 15, when 298 of the 2,403 employees (12.4%) exceeded the 20-hour limit. Some individuals exceeded the limit during as many as 11 of the 14 weeks, while others worked nine consecutive weeks in excess of 20 hours overtime. There were 114 instances of employees working 40 or more hours overtime in addition to 40 standard work hours, including two employees who each worked a total of 96 hours one week in January before the accident. 11 One of these individuals, a mechanical technician team leader, worked 60, 96.5, 94, and 80.8 hours per week in succession during the four weeks ending January 31, 1986.

 

Individual Workers

These levels of individual weekly overtime prompted a more focused analysis of the daily work hours for a selected sample of high overtime contract employees. The analysis was based on those time periods when the percent of overtime was unusually high and when processing of the Challenger Orbiter and Solid Rocket Boosters was taking place. The selection criteria were designed to include both those who worked several weeks of over time in succession and those who worked an unusually high amount of overtime during any one week. The sample for Thiokol included all employees, a total of 48, who met the following criteria during the six weeks from November 3 to December 15, 1985: 12

The Lockheed sample was based on more restrictive criteria in order to keep its size reasonable despite the much larger number of total employees:

These criteria were met by 239 Lockheed employees during the 13 weeks beginning October 26, 1985, and ending January 24, 1986. A final sample of 93 individuals was then assembled by combining all the workers who met the criteria in the three departments with the most cases. 13 Because a department's workload varies considerably each week depending on the current stage of Shuttle processing, each group of employees was examined only during the periods of peak overtime pressure for that department. These varied from seven to ten weeks extending from November 2, 1985 to January 24, 1986.

Three independent contributors to excessive overtime required examination:

The first two consequences were addressed together in the group analysis which follows. The third appeared to occur less extensively than the others and was therefore addressed separately on an individual case basis. Due to the sampling procedure, the results of the analysis reflect the upper extreme of the shiftwork continuum at Kennedy.

For the combined analysis the number of consecutive workdays was tallied separately for each group. The results are summarized in the following table. The mean number of consecutive workdays was calculated for all days, standard and overtime, and for overtime days only without any intervening standard eight-hour days or days off. The overtime results were further divided into increasingly longer workdays by two-hour intervals to better reflect the combination of working long hours several days in a row. The frequencies of occurrence of consecutive overtime days (i.e., two days or more) are also listed for each sample and overtime condition along with the maximum number of consecutive days worked by any one individual.

The left side of the table indicates that during this time both the Lockheed and Thiokol employees worked an average of seven to eight days in a row before a day off. Although only one employee in each group worked the highest number of consecutive days, 50 and 26 respectively, there were a considerable number who worked two weeks or more without a day off, especially at Lockheed. This distribution is readily apparent in Figure I which shows how often each company's employees worked more than six consecutive days. Separate frequencies of occurrence are plotted for increasingly longer work "weeks." The sharp drop-off seen in the Thiokol data after 12 days coincides with that company's policy limiting consecutive workdays to 13 or fewer and may reflect the effectiveness of that policy.

The right side of the table reveals the exposure of both groups to successive overtime workdays. The average number of consecutive workdays was computed for three categories of total daily hours: exceeding eight hours, equaling 10 or more hours, or equaling 12 or more hours. Thus, each succeeding category represents a more difficult work schedule. Within each company it is apparent that as the schedule becomes more difficult there is only a limited decrease in the exposure to continuous overtime days. Both the mean and the highest number of days change relatively little as the categories increase from 8 to 12 hours. Although....

 

Number of Consecutive Workdays at Kennedy Space Center: Selected High Overtime Employees

ALL DAYS

OVERTIME DAYS ONLY

.

.

Consecutive Days

Consecutive Days

Consecutive Days

Consecutive Days

.

> 8 Hours

* 10 Hours

* 12 Hours

.

Mean

Max

Mean

Max

Freq

Mean

Max

Freq

Mean

Max

Freq

.

Lockheed (N = 93)

7.36

50

4.62

15

451

4.56

14

447

4.62

14

180

Thiokol (N = 48)

7.97

26

4.26

13

113

4.12

9

112

3.95

7

108

* [greater or equal to]

 


Figure 1. Incidence of more than six consecutive workdays among a group of Lockheed (N=93) and Thiokol (N=48) shiftworkers at Kennedy Space Center. Time period of sample varies from seven to ten weeks extending from October 26, 1985 to January 24, 1986.

Figure 1. Incidence of more than six consecutive workdays among a group of Lockheed (N=93) and Thiokol (N=48) shiftworkers at Kennedy Space Center. Time period of sample varies from seven to ten weeks extending from October 26, 1985 to January 24, 1986.

 

[G3] ...differences in the frequencies of occurrence cannot be compared between companies due to the difference in sample size, they do indicate that there is a substantial decrease among Lockheed employees when the amount of daily overtime meets or exceeds 12 hours. Therefore, when Lockheed employees worked 12 hours or more, they did so for as many consecutive days as when they worked fewer overtime hours, but they did it less often. While none of the Thiokol employees worked this long for more than seven days in a row, 21.5 percent of the 93 Lockheed employees did so for ten or more consecutive days, some more than once. Furthermore, all of these workers were concentrated in two of the three departments.

There are four individual work histories that illustrate how overextended contract employees became when workload increased at Kennedy. Dave Owen of Lockheed has stated that there are cases of high overtime, but that they occur only occasionally for critically skilled workers in contrast to the general attempt by management to balance the overtime and keep it equally divided among employees. 14

The first individual is the mechanical technician team leader who was cited earlier for working 60, 96.5, 94, and 80.8 hours per week during the four weeks in January 1986. While this work all began on the first shift at about 8:00 a.m., it often extended until midnight. Across 26 consecutive days he worked only two standard eight-hour days, the remainder consisted of five 16-hour days, 16 12-hour days, and three days of 15, 14, and 13 hours each.

The second employee, an electronic technician, worked the third shift beginning about midnight in another department. From November 5 to 27, he worked 23 consecutive days including 14 that were 11.5 hours long. Then, after two days off, this technician worked 44 days in a row including 29 11.5-hour days, seven 12-hour days, and two 15-hour days.

The third individual is a manager of the technical shops which "are responsible for the fabrication, modification, refurbishment and maintenance of Shuttle ground support systems, equipment and facilities." 15 Starting October 28 he worked 26 consecutive days on the second shift beginning at approximately 4:00 p.m., including 17 days that were 12 or more hours long. In January he continued to work on the same shift for another 26 consecutive days with 18 days each lasting 12 hours or more. Sixteen of these were m succession.

The final worker is a lead electrician responsible for supervising other technicians working on ground support electrical systems and controls. He worked more consecutive days than anyone else in the Lockheed sample. The most disturbing aspect of his work schedule is that all 50 days were on the third shift. This shift usually lasts from midnight to 8:00 a.m., but on 38 of these days he worked 11.5 to 12 hours until about noon.

These four cases are extreme but by no means isolated. They illustrate the frequent pattern at Kennedy of combining weeks of consecutive workdays with multiple strings of 11 - or 12-hour days. Research has shown that either of these factors alone produces worker fatigue, but that together they represent a potential threat to safety and worker effectiveness. 16 However, it should be pointed out, as indicated in the report (page 70), the Commission did conclude that the launch site activities, including assembly and preparation, were not a factor in the Challenger accident.

Other studies have demonstrated that night work and shift changes produce sleep loss and fatigue by disrupting workers' Circadian rhythms. 17 Therefore, the additional factor of night shift [G4] work would enhance the problem for those working for extended long workdays. One group of such workers are the Morton Thiokol employees who typically work 12-hour shifts, either 3:00 p.m. to 3:00 a.m. or 3:00 a.m. to 3:00 p.m., from two to seven days (mean = 4.5 days) in a row while performing the stacking of the Solid Rocket Boosters. Both these extended work schedules disrupt normal sleep patterns by starting or ending at about the usual midpoint of night sleep, thereby producing substantial sleep loss. The occurrence of lengthened workdays of 12 to 16 hours in the preceding four case histories would also disrupt sleep by interrupting the worker's adjustment to his current shift schedule.

This concern is related to the third important aspect of the shiftwork issue, working different shifts. Shiftworkers are known to often require a week or more to adjust to new shifts, especially when the old or new shift is a night shift. 18 Thus, the greatest adjustment difficulty would be in transitioning from the first or second to the third shift or from the third to the second or first shift. A review of all the individual daily records from Lockheed and Thiokol revealed only 26 shift changes among the combined 141 employees in the sample. Thirteen were transitions to or from the third shift. This number of shift changes is relatively few for the length of time under study and suggests that the shift change factor does not warrant serious concern. The only exception to this view is raised by several Lockheed employees who regularly worked the first shift, but who changed to the second or third shift for only a few days several times during the sample period. One of them worked the following schedule: Day 1 - 8:00 a.m. to 4:00 p.m., Day 2 - 4:00 p.m. to 2:00 a.m., and Day 3 - 8:00 a.m. to 10:00 p.m. Such patterns of long hours coupled with abrupt shift changes are likely to increase the potential for shiftwork errors. 19

In summary, the amount of overtime at Kennedy was often quite high during the time period examined. The 20-hour overtime limit was exceeded 480 times by Morton Thiokol employees and 2,512 times by Lockheed employees, with some individuals experiencing much higher levels than others. The current management policies do not appear to be effective in minimizing this problem. Interviews conducted by the Commission suggest that senior management believes that overtime is being effectively controlled at the working level. When asked what the quality control was to make sure that one or two people weren't working so many hours that they were "burning themselves out and becoming a safety hazard," Dave Owen replied that

"Normally in a high percentage of cases it's voluntary. The employees volunteer. I think it's management discretion. The supervisor and the manager of the organization have to monitor it. And the people are not bashful; they get tired.... You need to realize, if you can, that people are very, for the most part, they are very perceptive of who's working the overtime. They have overtime books, and they themselves try to keep that overtime balanced, because they don't want the guy next to them working all the overtime and they don't get it." 20

During the same interview both David Owen and Robert Sieck agreed that there is no system at Kennedy for monitoring overtime from the safety perspective. 21 All documentation is oriented toward financial accounting and is not readily amenable to human factors analysis. The development of some type of regular overtime audit procedure combined with more restrictive controls might contribute significantly to reducing the potential safety risks associated with the shiftwork, especially that engendered by unanticipated launch postponements. One such approach has recently been described in a Nuclear Regulatory Commission report which reviews the literature on shiftwork, safety and performance 22 Much of the overtime revealed in the preceding analysis of workers at Kennedy would be beyond that report's suggested limits for workers at nuclear powerplants, an equally critical industrial setting. It should be recognized, of course, that there is always a trade-off between the desire to keep on duty those personnel with the greatest expertise and the need to guard against the undesirable effects of fatigue. Furthermore, it is not known how much the stimulation and excitement immediately preceding a launch might overcome some of these effects; however, such compensation clearly has its limits.

 

Overseas Travel

Several of the departments at Kennedy are also required to provide support for contingency landings during Shuttle missions. The assigned individuals travel overseas to alternate landing sites in Africa or Spain. Typically they depart Kennedy several days before a scheduled launch and return soon after completion of the mission. The travel is usually by commercial aircraft across five time zones.

Since most of the travelers are shiftworkers, the issue arises as to whether the added impact of a multiple time zone flight could exacerbate any performance deficit resulting from shiftwork either at Kennedy prior to flight, at the destination or at Kennedy upon return. The proximity of shiftwork to the flight, the shift worked and the direction of flight could all be expected to influence the traveler's ability to perform his job at any given time. 23 Therefore, the daily work hours of the sample of 93 high overtime Lockheed employees were examined to determine whether such travel occurred in conjunction with high overtime or difficult shift schedules. Only travel in support of missions 61-C and 51-L was reviewed.

Of the three departments represented in the sample, only one provided contingency overseas support for those missions. Fourteen individuals performed the necessary travel. Some of these experienced considerable overtime just prior to departing Kennedy. For instance, one worked 107.5 hours the previous ten consecutive days, including six 11.5-hour days and one 12-hour day. He went on duty for the first shift the day after he arrived in Spain. Three other employees who traveled together to Casablanca on January 19, 1986, reported working 96, 79.5, and 79 hours, respectively, for the week ending January 24. Only 16 of those hours were due to travel. Four of the fourteen individuals supported two missions during December and January. One had to return to Dakar only five days after returning from his previous 12-day visit.

This rather limited review of contingency support travel suggests that more attention ought to be paid to the scheduling of such travel in relation to shiftwork and overtime both at Kennedy and overseas. While the likelihood of a contingency landing may be remote, the performance requirements of the ground crew at the alternate site will be critical if the Orbiter experiences the type of emergency that dictates such landings.

 

Workload and Human Engineering

Any discussion about the safety implications of shiftwork must also address the situational aspects of human error in the workplace. Industry today is well aware that it is important to minimize the potential for such error in the design of industrial equipment and procedures. This requirement becomes even more critical when individuals are required to perform in a high technology environment with the added demands of shiftwork.

Shiftworker fatigue, high workload and faulty equipment design are a combination that can produce unnecessarily high safety risks. The previously described incident during the aborted launch of mission 61-C demonstrates the seriousness of such risks for Shuttle operations. In addition to operator fatigue, the Lockheed incident report on the dumping of 18,000 pounds of liquid oxygen specifically cites several other contributing factors related to the design of the console monitor display, inadequate operator training, poor communications and inadequate staffing. 24 While it may be expedient to address such problems as they occur and solve the problem at hand, the past record suggests that a thorough review of these human factors considerations is needed at Kennedy. The generic issues need to be addressed on a systematic [G5] basis to minimize the future potential for human error in this critical operational environment.

 

Managers at Risk: The Human Factors of Launch Pressures

The teleconference between Marshall Space Flight Center, Kennedy Space Center and Morton Thiokol, Incorporated, at Wasatch, Utah, focused attention on the advisability of launching mission 51-L and provided a clear opportunity for postponement. The decision should have been based on engineering judgments. However, other factors may have impeded or prevented effective communication and exchange of information.

One factor which may have contributed significantly to the atmosphere of the teleconference at Marshall is the effect on managers of several days of irregular working hours and insufficient sleep.

The extent of sleep loss was documented by Commission investigators who conducted interviews with the teleconference participants at Marshall and Kennedy specifically to reconstruct their daily activities during the five days preceding the accident. As shown in Figure 2, these interviews revealed that because of the launch scrub on January 27 certain key managers obtained only minimal sleep the night before the teleconference or had arisen so early in the morning that they had been awake and on duty for extended periods. 25

Time pressure, particularly that caused by launch scrubs and rapid turnarounds, increases the potential for sleep loss and judgment errors. This could be minimized by preventing launch support personnel, particularly managers, from combining launch support duty with office work on the same day. The willingness of NASA employees in general to work excessive hours, while admirable, raises serious questions when it jeopardizes job performance, particularly when critical management decisions are at stake.

 


Figure 2. Sleep patterns of Marshall Space Center telecon participants.

Figure 2. Sleep patterns of Marshall Space Center telecon participants.

 

REFERENCES

 

1. Report, "How Much Does Overtime Really Cost," Mechanical Contractors Association of America, Bulletin No. 18A, J January 1968.
Report, "Scheduled Overtime Effect on Construction Projects," A Construction Industry Cost Effectiveness Task Force
Report, The Business Roundtable, New York, November 1980.
 
Report, "Recommendations for NRC Policy on Shift Scheduling and Overtime at Nuclear Power Plants," U. S. Nuclear Regulatory Commission Report NUREG/CR-4248, PNL-5435, July 1985.

2. Report, "Incident/Error Review Board Items, FY84, FY85, and FY86", Lockheed Space Operations Company, KSC, 1984, 1985, and 1986.

3. Commission Interview Transcript, Sieck, R., and Owen, D.L., March 4, 1986, KSC, pages 27-28.

4. Commission Interview Transcript, Sieck, R. and Owen, D.L., March 4, 1986, KSC, page 30.

5. Report, "Investigation Summary of STS-32 LOX Replenish Valve Scrub", Lockheed Space Operations Company, KSC, February 19, 1986.

6. Memo, A. Aldrich to JSC Distribution, Subject: STS 61-C Launch, January 14, 1986.

7. Chart, "Overtime Percent by Second Level Directorate", Lockheed Space Operations Company, Program Management Systems, KSC, February 21, 1986.

8. Commission Interview Transcript, Sieck, R., and Owen, D., March 4, 1986, KSC, pages 4 and 8.

9. Morton Thiokol Inc. Management Policy No. 3130-22-01, "Working Hours," paragraph 14, August 7, 1985.

 
Morton Thiokol Inc. Management Policy No. 3140-22-01, ;'Recording and Paying Overtime/Extended Workweek," paragraph 4, August 22, 1985.
 
Lockheed Space Operations Co. Management Memo No. 013, from D.L. Owen, June 10, 1985.
 
Lockheed Space Operations Co. Management Directive V.6, revision 1, from A.R. Schroter, June 6, 1984.
 
[G6] 10. Chart, "Morton Thiokol Inc. Weekly Overtime Analysis for All KSC Employees (Oct. 6, 1985-Feb. 2, 1986)," Morton Thiokol Inc., KSC, April 1986.
 
11. Report, "LDRRPI Ad-Hoc Weekly Labor Hours by Employee/Department Report (Oct. 4, 1985-Jan. 31, 1986)," Lockheed Space Operations Company, KSC, April 28, 1986.
 
12. Chart, "Morton Thiokol Inc. Daily Work Time Analysis for Selected SRB Technicians and Inspectors at KSC (Nov. 3-Dec. 15, 1985)," Morton Thiokol Inc., KSC, May 6, 1986.
 
13. Chart, "Lockheed Space Operations Co. Daily Labor Data for Selected KSC Employees with High Weekly Overtime (Oct. 26, 1985-Jan. 24, 1986)," Lockheed Space Operations Company, KSC, May 9, 1986. An additional nine workers from these departments were identified after the study was concluded.
 
14. Commission Interview Transcript, Sieck, R., and Owen, D.L., March 4, 1986, KSC, page 9.
 
15. Job Description, "Statement of Functions and Responsibilities, Technical Shops Manager, Organization 33-41," Lockheed Space Operations Company, KSC, December 12, 1984, page 1.
 
16. Report, "How Much Does Overtime Really Cost," Mechanical Contractors Association of America, Bulletin No. 18A, January 1968.
 
Report, "Scheduled Overtime Effect on Construction Projects," A Construction Industry Cost Effectiveness Task Force Report
 
The Business Roundtable, New York, November 1980.
 
Report, "Recommendations for NRC Policy on Shift Scheduling and Overtime at Nuclear Power Plants," U. S. Nuclear Regulatory Commission Report NUREG/CR-4248, PNL-5435, July 1985.
 
17. Akerstedt, T., "Adjustment of Physiological Circadian Rhythms and the Sleep-Wake Cycle to Shiftwork." In S. Folkard and T.H. Monk (Eds.), Hours of Work, New York: Wiley, 1985, pages 185-197.

18. Akerstedt, T., "Adjustment of Physiological Circadian Rhythms and the Sleep-Wake Cycle to Shiftwork." In S. Folkard and T.H. Monk (Eds.), Hours of Work, New York: Wiley, 1985, pages 185-197.

19. Monk, T.H., and Folkard, S., "Shiftwork and Performance". In S. Folkard and T.H. Monk (Eds.), Hours of Work, New York: Wiley, 1985, pages 239-252.

20. Commission Interview Transcript, Sieck, R., and Owen, D., March 4, 1986, KSC, pages 12-13.

21. Commission Interview Transcript, Sieck, R., and Owen, D.L., March 4, 1986, KSC, pages 13 and 22.

22. Report, "Recommendations for NRC Policy on Shift Scheduling and Overtime at Nuclear Power Plants," U.S. Nuclear Regulatory Commission Report NUREG/CR-4248, PNL-5435, July 1985.

23. Graeber, R.C., "Alterations in Performance Following Rapid Transmeridian Flight." In F.M. Brown and R.C. Graeber (Eds.), Rhythmic Aspect of Behavior, Hillsdale and London: Lawrence Erlbaum Associates, 1982, pages 173-212.

24. Report, Investigation Summary of STS-32 LOX Replenish Valve Scrub," Lockheed Space Operations Company, KSC, February 19, 1986.

25. Commission Memorandum, "Human Factors Interviews at MSFC," Graeber, R.C., April 22, 1986.


Appendix F | Volume 2 Index | Appendix H