Hashem

Recent surveys reveal that design-build project delivery continues to grow in popularity as it overpowers other traditional methods. Research has revealed that, among other advantages, shortening project duration is one main reason why owners select design-build. But for design-build to maintain its current success, the collaborative system has to live up to owners’ expectations and continue to surprise the world with its ability to deliver projects with unbeatable speed.

An investigation of how planners at major design-build firms accelerate schedules yielded surprising and to some extent disappointing findings. While the planners realized the importance of overlapping design and construction activities and attempted to achieve it, very few of them followed a systematic, scientific method. The vast majority depend only on their own judgment and experience and input from other team members. This lack of a scientific basis for design and construction overlap led in many cases to costly and time-consuming rework and sometimes to significant schedule overruns.

Overlapping is good, but the question is how to best achieve it. Too much overlapping of design and construction can cause work to stop until the design catches up, or even worse, to demolition of completed work. But an insufficient amount undermines the benefits of design-build. Fast-track design-build project planners need a system supported by academia.

One answer may be my SAFEDB Methodology, which establishes a scientific basis to the overlapping process. SAFEDB is an abbreviation for ‘A Sound Approach to Formulate and Evaluate planning for large civil engineering fast-track Design-Build projects.’ The core philosophy is that in-depth knowledge of the nature and characteristics of each major project component is the key to success. Planning therefore has to follow scientifically based means and methods and must be proactive.

In SAFEDB, the planning process should start with the up-front selection of design solutions and include the adoption of the construction means, methods and techniques. The project plan therefore needs to be developed by a planning team of senior design, construction and scheduling professionals from within the design-build entity, rather than just by senior planners with input from design and construction professionals, as is common. By doing so, there is greater intensity, knowledge and thought poured into the schedule earlier in the process and a greater commitment to implement the schedule when the project is under way.

Setting Goals

SAFEDB has three basic planning goals: 1) to ensure that the selected design and construction solutions are the best to serve their intended purposes; 2) to foresee probable problems (failed tests, etc.) and to accommodate them into the schedule; and 3) to compress the schedule by the maximum safe, practicable application of overlapping activities.

These goals are accomplished by following three basic methodology components: 1) axiomatic design, 2) Graphical Evaluation and Review Technique (GERT), and 3) concurrent engineering.

Axiomatic design theory primarily is concerned with determining whether a design or construction solution is good or bad. It divides the design-build process into three main areas, or domains–functional requirements, design parameters and construction variables. Predecessor domains describe "what needs to be achieved", while successors describe "how it can be done."

The process starts by specifying the general functional requirements of the project in the functional domain. Then, the design domain responds with several design options, which then are measured against the functional requirements for the best fit. After the best design option is selected, the next level of functional requirements is laid out and the cycle continues until design is finalized.

The same process is applied to construction by zigzagging back and forth between design parameters and construction variables until construction also is finalized.

The GERT method is a further development of widely used Program Evaluation and Review Technique (PERT)–a network model developed in the 1950s that allows for randomness in activity completion times. GERT, developed in the 1960s, can handle more complex modeling situations. It explicitly defines the rework loops and branches for different outcomes in the project schedule and allows project planners to assign probabilities to their occurrence. Probabilities are estimated on the basis of historical statistics and experience with similar circumstances.

The GERT approach can accommodate the different scenarios that project activ-ities might encounter so that the project planners can decide the final resource allocation and/or countermeasure techniques.

Having selected the best design and construction solutions and accounted for potential branching and looping in SAFEDB steps one and two, applying the concurrent engineering theory is a decisive step toward optimizing the schedule by overlapping upstream design activities and pertinent downstream construction.

The process starts with establishing the fast-tracking characteristics of the various overlapped design and construction activities. That is basically defined in terms of the upstream activity evolution (UAE) and the downstream activity sensitivity (DAS). UAE is the rate of progress in the development of the upstream design activity, while DAS describes how sensitive the downstream construction activity is to changes in information transferred from the upstream design activity.

In general, there are four basic modes of combination between UAEs and DASs: Slow UAE with Low DAS, Fast UAE with Low DAS, Slow UAE with High DAS and Fast UAE with High DAS.

The SAFEDB methodology has established that overlapping of activities should be carried out only after careful consideration of the pre-defined activity’s fast-tracking characteristics. The table below summarizes the outcome of the portion of the SAFEDB methodology looking into overlapping. It provides a recommended generic numerical overlapping strategy comprising all possible combinations of upstream and downstream activity fast-tracking characteristics. The values shown indicate the percentage of the upstream design activity duration at which downstream construction activity may start, in a start-to-start mode. Percentages of actual completion of the upstream design activities likely to have been accomplished at the time the downstream construction activity starts also are given in red. Click here to view pdf chart

Reliability of the information developed in the upstream design activity has a major effect on the extent of the safe overlapping of the same with the downstream construction activity. In a design-build environment with a reliable upstream design activity, the work produced can be readily passed along to initiate the downstream construction activity, resulting in a speedier design-build process.

One example would be the foundation design for a multi-story building. Column loads can be calculated first, once the structural model is established and decisive load combinations are examined. Foundation design then can be finalized and released for construction while design of other structural elements (columns, slabs, beams) proceeds in parallel.

On the other hand, if the upstream design activity is highly unreliable, overlapping should be avoided to minimize the possibility of reworking in the downstream. An example of this would be hydraulic design of a closed-loop pressurized water system. Hydraulic design should be fully finalized after examining the various potential pumping and shut-down scenarios before releasing any part of the network for construction.

Sensitivity of the downstream construction activity to changes or errors in the information transferred from the upstream design activity also has a similar effect on the amount of overlapping. For certain scenarios in which a highly unreliable upstream design activity is followed by a highly sensitive downstream activity, a schedule buffer, such as a finish-to-start lag, is recommended. In this overlapping strategy, the maximum amount of time allowed for upstream activity completion is set at up to 160%. By doing so, planners will have enough time to discover potential errors in the upstream design activity and fix them, thereby avoiding any costly impact of these errors on the downstream construction activity.

For the most effective implementation of the SAFEDB methodology, planners and project managers should focus on major activities lying on the project’s critical path. This should be done until they move the project off that critical path and onto a new critical path as overlapping continues. The process continues until no further overlapping is practical or feasible.

Tools Needed

Unfortunately, the current state-of-art of the commercial computer-aided planning programs do not yet support the SAFEDB methodology and it can only be applied manually at this stage. This could be an exhaustive task if applied to many parts of a project. Further development of the currently available project planning computer programs and software (such as Primavera P3) to accommodate design-build needs to be based on SAFEDB principles and could be a major advancement of design-build projects. That could be done by expanding the scope of activity definition to include fast-tracking parameters and to account for looping and rework cycles in a probabilistic fashion.

Sherif Fouad Hashem is head of the project management department,
Consulting Engineering Group, Doha, Qatar. He can be reached at sheriffh@link.net or cell phone 974-581-7030.