Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
8 Project Cycle Time Reduction INTRODUCTION The NRC report on DOE project management (NRC, 1999, p. 89) observed that "DOE WM (now called EM) projects took an average of three times longer to complete than comparable projects in industry and other government agencies." This observation was based on data from the mid-199Os because up-to-date infor- mation on project durations was not available. It was often unclear just when a project was actually started or when it was finished (put into operation). Never- theless, there seems to be general consensus that most DOE projects could, and should, take substantially less time to complete. PREREQUISITES FOR REDUCING CYCLE TIME To reduce project cycle times, it is necessary that project management processes and procedures be consistent and effective. Previous NRC reports stated that the DOE project development process has not been in control of cost, schedule, or performance metrics; however, steps are currently being taken that are intended to bring the process under control (NRC, 2001a, 2001b, 2002a). Only when most projects are dependably under control will it really be possible to reduce process cycle times. Proposed projects are justified on the basis of estimated budgets and sched- ules, and virtually all are claimed to be essential to the DOE mission, but project execution plans and critical decision (CD-0) submissions do not explicitly answer the question, How much does the project's value (to DOE and to the public) 49
so PROGRESS IN IMPROVING PROJECT MANAGEMENT AT THE DOE change if it is completed earlier or later? To know whether it is beneficial to complete a given project earlier, it is necessary to estimate the net value (not the cost) of the benefits to be obtained from the project, and then to estimate the net value if the project is delivered 1 month or 1 year earlier. Earlier delivery of projects is not an end in itself but is a means of reducing the time needed to obtain the benefits conveyed by the project and benefits obtained sooner are more valuable. This determination requires a clear distinction between value and cost. An essential step in reduction of project cycle times is to assign higher priorities to those projects whose earlier delivery would be cost-effective, based on the net increase in value to the government attributable to the earlier receipt of benefits less the increase in cost, if any. However, the committee observed that no such cost-benefit analysis has been carried out for DOE projects. The fact that it is difficult to quantify project benefits is not regarded as an acceptable reason for not doing so. Completing projects earlier requires identifying projects that are needed sooner, which means setting priorities. Decisions on project delivery schedules will be valid insofar as they are based on objective, unbiased information. If there were clear strategic plans for all the DOE functions, and if the DOE implemented a portfolio planning and management process, program manage- ment would be much improved. The process would be much easier: Each set of project proponents could then match its project's attributes against the known mission requirements and planned project portfolio. Proposed projects with low mission need could be deferred or terminated, and those deemed critical to mission should be advanced. This prioritization and reduction of cycle time would lead to projects completed sooner and fewer projects in the pipeline. Absent a function- ing portfolio management process, project proponents and DOE managers cannot reliably determine which projects are considered to be mission critical by DOE headquarters, so they propose more projects than can or should be funded. This situation generates a requirement for senior DOE management review points (critical decision points) to determine which proposals are most valuable with respect to strategic mission requirements. The less clear the mission statement, the more management needs to be involved in the decision process. EFFECT OF DOE ORDER O 413.3 ON PROJECT CYCLE TIME The committee has heard from some DOE and M&O contractor personnel that the critical decision points set up in DOE Order O 413.3 have greatly increased project delays. The committee knows of no valid data confirming whether or not O 413.3 delays projects, as no projects have gone completely through the O 413.3 process. Even if such data were available, the state of the DOE historical project database is such that no statistically valid comparisons could be made. Projects that are now in construction were started before O 413.3, and those that were started under O 413.3 are only at CD-1 to CD-2. If the review process leads to projects that are completed successfully, on time, and on budget
PROJECT CYCLE TIME REDUCTION 51 and the absence of a review process leads to projects that overrun the schedule and budget, then the claim that project reviews delay projects becomes moot. DOE is interested in outcomes, but the outcome of interest is not the length of the review but the success of the project (DOE, 2000~. Therefore, claims that O 413.3 decision points are delaying projects are unsubstantiated. If critical decision reviews take longer than deemed necessary, then DOE senior management should fix the management review process, not eliminate management reviews. Senior management reviews actually add value to the process, even if they only weed out projects that should not proceed. As noted above, if strategic directions and mission requirements are not clearly enunciated, it is the legitimate role (in fact, the responsibility) of senior management to determine which of the proposed projects are most aligned with the strategic goals of the organization. The views of successful project managers in industry on this subject were dis- cussed at the November 2001 government/industry forum on the owner's role (NRC, 2002b). Making strategic decisions is what senior managers in any organization, public or private, are required to do. Senior management can delegate this authority when they have a high degree of confidence that lower echelons will replicate decisions of senior management. Absent this confidence, senior man- agers will have to make all strategic decisions. Given that big projects are the core of DOE's mission, senior-level management would be remiss not to be involved in critical decisions. The committee believes that authority to decide which projects will be funded should in no case be delegated to M&O or M&I contractors. DOE Order O 413.3 essentially defines five critical decision points (CD-0 to CD-4. This does not seem to be an excessive number compared with private industry, which often requires more senior management decision points and con- siderably more intense senior management involvement in projects of far less dollar value than DOE projects (DOE, 2000; NRC, 2002b). What was remarkable about DOE projects in the past was their micromanagement by many midlevel managers at headquarters, combined with a relative lack of attention at high levels, even for billion-dollar projects authorized as essential to the national defense (e.g., the National Ignition Facility, prior to 1999~. In the absence of reliable data, the committee tried to explain why manage- ment review at the critical decision points takes such a long time: The committee has observed that critical decision reviews often include polished PowerPoint presentations and even rehearsal sessions prior to the actual executive review. It takes time to prepare these presentations and get them reviewed and approved by managers up the line even before they reach the ultimate decision maker. As a result, although these iterative reviews may serve a vital communications role, and the need for concurrence and political buy-in at various levels may justify this process, they strengthen the perception that the process is cumbersome and too formal.
52 PROGRESS IN IMPROVING PROJECT MANAGEMENT AT THE DOE One reason for this formality may be the general state of communication within DOE. The committee has observed a generally low degree of communica- tion both between programs and within programs. If senior DOE managers were well informed about projects, it would not be necessary for the reviews to cover the entire project territory in such formalized briefings. The responsibility for ensur- ing that senior managers are well informed about projects lies both with the senior managers themselves and with the project and functional managers below them. ELIMINATING DELAYS FROM CRITICAL DECISIONS There are at least three ways that critical decision reviews can delay a project: . Materials (acquisition plans, risk management plans, etc.) are inadequately prepared and are sent back for rework. · Materials are adequate but the project is deemed unnecessary or of low priority and does not pass the critical review, and it is either terminated or recycled for further work. · The preparations for the critical decision review are on the project critical path. Clearly, decisions are on the critical path they are intended to be. But preparations for reviews need not be on the critical path. The requirements of O 413.3 are known. Therefore, a well-managed project should be planned such that adequate project justification and documentation, including acquisition plans and risk management plans, are standard parts of preproject planning, so that the preparations for project review presentations are not on the critical path. Project justifications are almost universally based on the premise that the proposed project is the only possibility, exactly as it is proposed, and that the only alternative to this specific project is no project. A single alternative may be presented despite the fact that the project scope and specifications were undoubt- edly the result of negotiations and compromises among all project proponents and participants. This means that if the decision maker does not like either alternative yes or no the justification has to go back for rework. Inadequate preproject planning is at fault here; front-end project planning documents should address the following: . Alternatives to the proposed project. If this exact project is not approved as defined, then there should be an alternative ("Plan B") for meeting mission requirements. Issues, capabilities, andfeatures not included in the proposed scope but capable of being added later on. These might include expanding capacity, acquir- ing more or newer instruments, upgrading capabilities, etc. If this information is on hand, the decision maker might choose to bring some of these features forward into the proposed initial project scope. .
PROJECT CYCLE TIME REDUCTION 53 · Capabilities that are precluded by the proposed project scope, project architecture, or technology. These are the opportunity costs: the things that one could not do (or could do but at prohibitive expense), now or in the future, if one were to adopt the proposed solution. Given this information, the decision maker might find that some essential or desired future capability has been ruled out by the proposed solution and therefore might require a different solution. The failure to provide this kind of essential information has been character- istic of the DOE culture. Some project proponents told the committee that they do not trust DOE senior management to make the correct decisions, even when they have the essential information. This lack of trust in senior managers leads the senior managers, in turn, to not trust project proponents to brief them thor- oughly or make the best decisions for the agency independently, whereupon they insist on making all the decisions themselves. COMPLEXITY INCREASES THE RISK OF DELAYS The nature of the DOE project development process encourages many projects to grow by accreting more scope and functions and by gaining internal or external political support until they become megaprojects. Adding more func- tions or scope increases complexity, and the difficulty of coordinating these multiple functions increases exponentially with complexity (Merrow et al., 1988~. Flyvbj erg documents overwhelming evidence that the costs, durations, and risks of public works megaprojects are consistently underestimated, and DOE projects are apparently not immune from this syndrome (Flyvbj era et al., 2002~. Some DOE capital acquisition projects are at the same time R&D projects. The reason for these dual objectives is said to be speed. Project proponents feel that they do not have time to research and refine the technology from bench scale to full scale before embarking on the full project, so the R&Dis done on the full- scale capital acquisition project. Consequently, the projects are delayed for technology development and take longer than if the necessary R&D had been competed first and the costs are much higher. Technology development or R&D on a capital project provides a perfect excuse for delays and cost overruns, even when they may actually be due to other problems. In any case, if one accepts the proposition that major systems acquisition and R&D must be conducted concur- rently, the R&D should not be on the critical path, and critical path activities should be buffered against the results of the R&D process. If there are technol- ogy development risks, the risks should be concentrated, not diffused throughout the project, where they will be more difficult to manage and control. If there are technical risks with system components, final system integration should not be a risk factor, as system integration and startup will certainly be on the critical path and will most likely delay the project, especially if system integration requires
54 PROGRESS IN IMPROVING PROJECT MANAGEMENT AT THE DOE rework of components to achieve compatibility. (See the discussion of first-of-a- kind projects in Chapter 7.) REDUCING PROJECT TIME AND SUPPORTIVE MANAGEMENT CULTURE Earlier project completion requires more flexibility on the part of all partici- pants. DOE's ability to adjust rapidly to circumstances is limited partly by the Federal Acquisition Regulations (FAR) but even more by DOE practices. To an outside observer, a faster project process appears even less well organized than projects on normal timetables. Reduction of project cycle time depends on effec- tive decentralized decision making. This is exactly why the project delivery process should be brought under control before attempting to deliver projects earlier. An uncontrolled process at normal speed will be worse at high speed. To perform projects faster, a high degree of trust is necessary between participants at all levels and locations. Such cohesion, which comes from a spirit of participa- tion in a common enterprise to achieve common goals, does not always prevail at DOE or its contractors. STAFFING FOR REDUCED PROJECT CYCLE TIME Delivering projects early requires more dependence on good project manage- ment. DOE's deficiencies in project manager training, career paths, professional development, and related issues have been documented elsewhere. To deliver projects early and on budget, project managers must have authority and senior management support and must know how to use them to remove obstacles to project progress. DOE often assigns world-class scientists and technical experts as project team leaders; however, this can be counterproductive. In general, engineering managers, or technical leaders, should be distinct from project man- agers. The function of a project manager should be to keep the project on scope, on budget, and on schedule. One place in which time could be cut from projects is the front end. This does not mean reducing time spent in project definition, preproject planning, or conceptual engineering, which the committee has emphasized in previous reports (NRC, 2001b), but reducing the time wasted getting organized and time delays due to understaffing at the front end. Project execution plans are essential and their preparation does not delay projects unless there are delays in setting up project organizations, staffing integrated project teams, making timely decisions, all of which should be avoidable. Time lost to delays here is very difficult to recover later on, and even if recovery is possible, it can be very expensive. The true cost of the front-end process is the cost (that is, the value foregone) of slow progress and later project delivery, not just the cost of staff to plan, manage, and engineer the project. Front-end planning activities should have resources assigned
PROJECT CYCLE TIME REDUCTION 55 on a timely basis and explicit milestones or deadlines. Many planning and engineering functions can be accomplished concurrently or in an overlapping manner. One reason many projects start slowly is that personnel are not made available from other assignments. A project has not really begun unless a project manager and an integrated project team are actually at work. There is hardly a case in which a project team was assigned too early. Early assignment of project teams is necessary to start front-end planning. Even though projects with shorter cycle times require more flexibility in the organization, they also, perhaps para- doxically, require more planning than business-as-usual projects. The planning does not necessarily entail making all decisions early; it may, however, entail more advanced planning of when decisions must be made and providing the information needed to make them. FINDINGS AND RECOMMENDATIONS Finding: Undersecretary Robert Card has enunciated a new strategy for Environ- mental Management (EM) that stresses earlier completion of site cleanup and remediation and earlier closure of sites or their turnover to private industry. The EM organization is reorganizing to fulfill this new strategy. Although it appears that much of the time reduction will be due to a reevaluation of the necessary end states, which may involve negotiations with stakeholders, the committee considers this initiative an important step toward DOE controlling its projects rather than being controlled by them, as has been the case. To make progress, it is necessary to believe that projects can be controlled and delivered earlier rather than believing that nothing can be done and that the process will require 70 years to complete. It is too early to determine whether the new EM organization will be successful, but the committee considers active attempts to get projects under control, to define strategic directions, and to align projects with strategy to be superior to passivity. Recommendation: The strategy of achieving earlier completion of site remediation and closure or turnover of sites should, if successful, reduce environ- mental risks substantially and save U.S. taxpayers many billions of dollars. This initiative should be supported and continued. Recommendation: In addition to redefining end states, DOE EM should con- sider all possible methods for improving its project management processes, pre- paring its project managers, and achieving earlier project completion, some of which are outlined above. Recommendation: Program offices in DOE other than EM should also consider opportunities for earlier project delivery through application and implementation of the principles cited above.
56 PROGRESS IN IMPROVING PROJECT MANAGEMENT AT THE DOE REFERENCES DOE (Department of Energy). 2000. Program and Project Management for the Acquisition of Capital Assets, Order O 413.3. Washington, D.C.: Department of Energy. Flyvbjerg, B., M.S. Holm, and S. Buhl. 2002. "Underestimating costs in public works projects: Error or lie?" Journal of the American Planning Association, 68(3): 279-295. Merrow, Edward W., L. McDonnell, and R. Yilmaz Arguden. 1988. Understanding the Outcomes of Megaprojects: A Quantitative Analysis of Very Large Civilian Projects. Santa Monica, Calif.: The Rand Corporation. NRC (National Research Council). 1999. Improving Project Management in the Department of Energy. Washington, D.C.: National Academy Press. NRC. 2001a. Improved Project Management in the Department of Energy. Letter report, January. Washington, D.C.: National Academy Press. NRC. 2001b. Progress in Improving Project Management at the Department of Energy, 2001 Assess- ment. Washington, D.C.: National Academy Press. NRC. 2002a. Progress in Improving Project Management at the Department of Energy, 2002 Interim Assessment. Letter report, May. Washington, D.C.: National Academy Press. NRC.2002b. Proceedings of the Government/Industry Forum: The Owner's Role in Project Manage- ment and Preproject Planning. Washington, D.C.: National Academy Press.
