The National Nuclear Security Agency’s (NNSA’s) stated mission is to “enhance global security through nuclear deterrence, non-proliferation, counter-terrorism, naval nuclear propulsion, and national leadership in science, technology, and engineering.”1 At the very core of the mission responsibility of the LLNL, LANL, and SNL is the maintenance of a safe, secure and effective nuclear arsenal. In the post-Cold War period—in the absence of nuclear-explosion testing and the production of new weapons—the responsibility to maintain an enduring stockpile requires advanced science, technology and engineering competencies.
NNSA headquarters elements provide “policy, priority, and program funding guidance, along with oversight and programs toward defined strategic goals.”2 NNSA Site Offices are located in proximity to the laboratories to provide “direct budget, regulatory and contract oversight, and administrative authority for these laboratories.”3 The NNSA management approach seeks to integrate “leadership, people, and processes to better accomplish [the] goals of a unified National Security Enterprise.”4
This approach has resulted in an increased centralization of science and technology planning and direction, in which the laboratories have lost some of their historic independence and self-initiative, and which has resulted in top-down tasking to the laboratories. For example, in the weapons area, detailed surveillance and life extension programs of specific weapons systems dictate which of the laboratories does what and when to maintain the safety, security and reliability of the stockpile. If left unchecked, this management approach increasingly takes initiative and control out of the hands of working scientists and engineers, and places it in less expert hands in Washington. While many NNSA officials are experienced scientists with relevant laboratory experience, their headquarters jobs remove them from day-to-day research activity.
A parallel trend has been toward a contractual relationship that is increasingly focused on nonscientific operational matters, such as security, safety, administration, facilities management, financial management, and other such functions. For example, in defining specific criteria for the determination of award fee and award term (in the Performance Evaluation Plans), more of the award depends on meeting operational goals than depends on meeting goals associated with the quality of science and engineering (and other mission-related goals)
The management relationship between the Department of Energy, NNSA, and its national security laboratories is defined by detailed contracts focused on assuring that the work of the laboratory is conducted in an environmentally responsible, safe and secure manner, and that operations of the laboratory maintain fiscal integrity. The current management contracts award substantial financial rewards for contract performance in these areas. At LANL and LLNL, roughly 30 percent of the fee is fixed, and 70 percent is at risk in yearly performance evaluations (6-10 percent of the at-risk fee is based
_______________________________
1 NNSA 2011 strategic plan.
2 NNSA 2011 strategic plan.
3 NNSA 2011 strategic plan.
4 NNSA 2011 strategic plan.
directly on quality of science and engineering at LANL and LLNL: Appendix D).5,6 This formula is designed to provide incentives for a high degree of management performance, which can be constructive in many environments. However, in an environment of broken trust, it carries a high risk that management will focus almost entirely on those contractual scoring criteria that account for the majority of the award fee, to the detriment of the science and engineering components of the mission.
A senior staff member at LANL provided some written comments to the study committee that captures the situation very well:
When I started as a young postdoc and then later in my career as university professor and also here at the laboratory, there was a social contract, which basically said ‘You will never get rich in science, but we treat you as adults, respect you for your commitment, and in turn you can pursue science and have fun.’ Today, this contract is badly broken … an atmosphere of distrust … rigorous control and checks.
How else could one explain the fact that today the signatures of [3-4 people] are required if I want to take my laptop home to work from home? I also need to write a half page justification why I want to work from home. If I want to attend the meeting of the division of nuclear physics of the APS, I need signature of [five people] … Where academic freedom once reigned … we have today a laboratory totally driven by risk averseness. We are drowning in paperwork and regulations. I know of three world-class scientists just in my group, who left … because they could not work in this environment anymore. Many more in other groups and divisions also left.
An LLNL employee with over three decades of experience explained the effect that this environment has on high-quality S&E:
I have seen our efficiency drop by at least a factor of two over the last two decades, and the inefficiency accelerated after the contract change from UC to LLNS. The laboratory is being micro-managed by DOE, and now the new contractor, to the detriment of this country. I worked hard, and I’m sometimes frustrated by the bureaucracy that does not have a long-term view of the lab. It seems that concern about risks overrides scientific progress constantly. Often times, I will not initiate or take on difficult R&D assignments because of the unfunded hoops I have to jump through …
An erosion of trust on both sides of the relationship shapes the oversight and operation of the laboratories. This in turn has resulted in excessive reliance on operational formality in important aspects of laboratory operations, including the conduct of science and engineering at the laboratories. Operational formality is the application of specific rules and predetermined procedures to the accomplishment of tasks. This approach derives from industrial practices, where it is often important to assure goals such as safety by specifying exactly how tasks are to be done and then taking measures to ensure that these steps are strictly followed.7 While the application of “follow the numbers” to ensure safety in selected tasks seems obvious, so does the mismatch of this approach to creative activities such as S&E. This erosion of the trust relationship is prominent with respect to LANL, where past failures attracted much national attention and public criticism. But it has also spilled over to LLNL and SNL, where management relationships also have acquired considerable operational formality.
While some laboratory S&E staff believes the excessive use of operational formality is a choice imposed by the M&O contractors, or by the contracts, the study committee did not see evidence of that.
_______________________________
5 At Sandia, where the at-risk fee is much smaller (although the fixed fee is roughly the same), the ratio is reversed.
6 For example, over $54 million of fee was at risk to performance assessments in FY2010 at LANL, of which $44 million was granted.
7 See, for example, Defense Nuclear Facilities Safety Board, Operational Formality for Department of Energy Nuclear Facilities and Activities, Technical Report DNFSB/TECH-15, March 1997.
When laboratory employees were questioned about heavy-handed bureaucratic processes, they could not point to their origin; that was true even for managers. The contracts and their incentives do not seem to encourage or mandate this. One senior SNL employee suggested that conservatism can accrete when there are layers of rules and processes, with little trust about who is going to take on risk.
Trust can be considered in two different ways: one concerning reliance, and the other confidence. Reliance means believing in the other party’s character and ability: can the other party be believed? Does the other party know what he/she is talking about? Do I have faith in the other party’s knowledge and expertise? Confidence means believing that I can depend on something in the future regarding another individual or group. Can I rely on the other person to do what they said they would do? Based on extensive discussions, the study committee thinks that if it were to ask NNSA, the laboratory managers, or the scientists and engineers at the laboratories these questions, none would answer in the affirmative. There is a persistent level of mistrust. While some progress has been made in recent years under current NNSA and laboratory leadership, much more is needed to repair the damage that has been done.
Finding 4.1. There is evidence of poor communications and lack of transparency at the highest levels, as illustrated by NNSA and laboratory leadership reporting significantly different assessments of the current management and operational relationship. The degradation of trust—whether confidence or reliance—is frequently accelerated in an environment of poor communication and lack of transparency. Discussions at study committee meetings indicated a persistent level of mistrust between NNSA staff and the laboratory scientists, particularly at LANL and LLNL.
Finding 4.2. The LANL and LLNL Site Offices are organized and staffed largely for monitoring compliance of the laboratories with extant DOE and other operational regulations. This reflects mistrust of laboratory management and staff to execute its mission responsibilities effectively and with reliable commitment to safety, security, and environmental concerns.
The study committee recognizes the responsibility to follow federal regulations about environment health, safety, and security, but also argues for a balanced approach that maximizes scientific flexibility within those requirements.
THE EFFECT OF OPERATIONAL FORMALITY ON EXPERIMENTAL WORK AT THE LABORATORIES
Experimental science is at the very heart of the scientific method, which relies on gathering empirical and measurable evidence subject to specific principles of reasoning tested through experimentation. Experimentation leads to discovery, and also provides essential validation for modeling and simulation.
The study committee observes that operational formality, which has been the by-product of the loss of trust in the laboratories’ ability to maintain fiscal integrity and the safety and security of its work, is not a good basis on which to conduct productive, creative experimental work. Its checklist-based methods are demonstrably valuable for high-risk tasks, but onerous when nimble thinking and innovation are required. S&E staff and some managers at all three laboratories told the study committee that experimentation is becoming more difficult to pursue, and therefore less common, because of burdensome steps that must be completed associated with purchasing, safety checks and certifications, and so on. Thus, there is already some evidence that science and engineering at the laboratories are relying less on experimentation, which has worrisome implications for the S&E.
Finding 4.3. Increasing operational formality contributes to a bias against experimental work. Without a strong experimental program, the quality of scientific and engineering at the laboratories will be at risk, as will the core mission of these laboratories.
NNSA needs to reexamine the roles and responsibilities of federal oversight officials and laboratory management, and a mechanism needs to be devised to resolve differences that occur in executing roles and responsibilities in laboratory operations and programs. Excellent science and engineering is at risk when laboratory scientists and engineers do not perceive that they are in a partnership that encourages them to bring forth their creative ideas to solve problems vital to our national security. In the broader science environment, such conflicts are typically settled through peer review and open discussion. Resolution through back channels sows mistrust. By the very nature of the laboratories’ mission, much of the work is done in a closed, classified environment. This adds complexity when trying to resolve scientific conflicts, but does not remove the necessity for doing so.
Successful partnerships, like successful societies, depend upon a high level of trust. Like barnacles on the bottom of a boat, mistrust accretes and accumulates over time until it compromises performance. Broken trust requires repair if the long-term performance of the laboratory missions is not to suffer. Due to the degree of mistrust that has encrusted over time, repairing that broken trust will require considerable time and effort. Mistrust is a highly stable phenomenon and can last for years if not decades. Therefore, attempting to fix things all at once and quickly is naïve and likely to fail.
Recommendation 4.1. The study committee recommends that NNSA and each of the laboratories commit to the goal of rebalancing the managerial and governance relationship to build in a higher level of trust in program execution and laboratory operations in general.
Recommendation 4.2. The study committee recommends that NNSA and the laboratories agree on a set of principles that clearly lay out the boundaries and roles of each management structure, and also that program managers at headquarters, the Site Offices, and in the laboratories be directed to abide by these principles.
For example, the site manager and the director and/or deputy director of each laboratory could establish, in consultation with other laboratory staff, a process to identify and agree on eliminating certain oversight procedures that are not necessary or related to the overall goals of the laboratory. Similarly, some mechanism could be established to filter program tasking at both the headquarters level and at the laboratory senior management level to assure that each tasking is necessary and consistent with the agreed management principles.
Recommendation 4.3. The study committee recommends that the goal of rebalancing the relationship and the set of principles laying out the boundaries and roles of each management structure be memorialized in memoranda of understanding between NNSA and its laboratories. Performance against these understandings should be assessed on an annual basis over a five-year period, and reported to Congress. 8
One sign of broken trust reported to the study is that mid-level issues were elevated to the laboratory director level because there was no clarity about how to resolve disputes between a laboratory and an NNSA Site Office. Another example was a recent instance in which NNSA HQ tried to overrule a laboratory’s best scientific judgment about how to carry out a scientific task. Subsequently, language appeared in a congressional report opposing that NNSA order.
_______________________________
8 The committee observes that it is important to design this approach to be self-correcting and to avoid problems such as: (1) adding to a check-list approach to management; (2) enforcing measures that annual assessment shows to be unworkable; and (3) requiring Congressional intervention when not needed.
Finding 4.4. There is no apparent mechanism by which the NNSA and its national security laboratories can negotiate a balance between competing policy, programmatic, and technical demands. In an environment that lacks trust, lack of an effective process for resolving such conflicts leads to situations that can be viewed either as NNSA inserting itself in an inappropriate operational role or the laboratories inappropriately challenging NNSA’s role.
A better mechanism could be established for resolving technical disputes, without elevating them to top NNSA management and congressional levels. A technical advisory committee, established at the NNSA level, would be a helpful mechanism for filling this gap in S&E management. More generally, such an advisory committee could monitor progress on other aspects of roles and responsibilities described in this chapter.
