Sunteți pe pagina 1din 7

DESIGN-BUILD ESTIMATING FROM THE BOTTOM UP

Paul W. Dutton Jacobs Associates John M. Stolz Jacobs Associates James J. Brady Lachel & Associates Carlton M. Ray District of Columbia Water and Sewer Authority ABSTRACT The design-build project delivery method presents unique challenges to the preparation of an engineers estimate and schedule, which are not encountered with traditional design-bid-build projects. This paper will present an overview of how a true bottom-up cost estimate with an integral schedule was developed as a tool to overcome the many estimating and scheduling obstacles presented by the District of Columbia Water and Sewer Authoritys (DC Water) Blue Plains Tunnel Project. The project consists of five shafts and approximately 7,300 m (24,000ft) of tunnel. This bottom-up estimating method has served to validate the existing budget and schedule, and because of its adaptability to the changing scope and design refinements, it has acted as an aid in setting realistic milestone dates, updating budget needs, and performing risk and sensitivity analyses. INTRODUCTION The District of Columbia Water and Sewer Authority (DC Water) provides waste water collection and treatment for the entire District of Columbia and parts of suburban Virginia and Maryland. DC Water is currently in the midst of a massive program now known as the DC Clean Rivers Project. This project is the result of an agreement (consent decree) with the United States Environmental Protection Agency (EPA) to improve water quality in the Districts waterways, specifically Rock Creek and the Anacostia and Potomac rivers, by implementing a long term control plan (LTCP) to reduce sewagecontaminated storm water discharges (CSOs). The need for the LTCP is a result of many older developed areas of the District having combined sewer systems that carry both sanitary and storm flows. During periods of heavy rainfall, the combined sewers are overwhelmed by the increased flows causing overflow of untreated sewage into these rivers and their tributaries. Project Overview The Blue Plains Tunnel (BPT) was not part of the Districts initial LTCP, but was added as a result of an amendment to the consent decree stemming from a separate effort aimed at reducing the amount of nitrogen in treated effluent discharged to tributaries of Chesapeake Bay. This effort applies to all areas served by DC Water, not just the District of Columbia, thus the BPT is funded from multiple sources. The BPT is the first major construction project to be executed in the implementation of the Districts LTCP, and it will ultimately serve as storage and also convey storm water to the Blue Plains Advanced Wastewater Treatment Plant (Blue Plains). The tunnel contract includes final 44

Design-Build Estimating from the Bottom Up 45

Figure1. Anacostia River projects in the facility plan design and construction of approximately 7,300 m (24,000ft) of soft ground tunnel with a segmental concrete lining, five large diameter shafts, a diversion structure, and several smaller elements that will link the tunnel to future infrastructure. Planning In accordance with the LTCP, DC Water consultants produced a Facility Plan based on several years of research and study. The Facility Plan specifies minimum program parameters, such as storage volumes and drop shaft locations, that will be required to meet the projects goals. In order to obtain a better value, DC Water divided the scope of work in the Facility Plan into multiple contract divisions (Figure1). The reasons for this division were to provide more manageable contracts, stimulate competition between contractors, and offer more opportunities to meet DC Waters fair share Minority and Women Owned Business Enterprise (MBE/WBE) objectives. DC Water also considered alternative contracting methods, a departure from its traditional designbid-build (DBB) approach. After much consideration and debate, DC Water decided to

46

Contracting Practices and Costs

proceed with the design-build (DB) contracting approach because of the potential cost and schedule savings. The BPT will be DC Waters first project to be constructed using this approach, and its effectiveness will be evaluated before using this method for subsequent large tunnel contracts that form the core of the LTCP. In March 2009, DC Water formed the Program Consultants Organization (PCO), led by Greeley and Hansen and Jacobs Associates, to provide more input on contract packaging, further detail planning, start design work, and oversee construction of these facilities. THE BOTTOM-UP ESTIMATE Project cost is of particular importance for the BPT because of this new contracting approach and its joint funding sources from the multiple jurisdictions served by DC Water. Anyone familiar with underground estimating knows that scheduling and estimating costs with certainty for tunnels is challenging, even with a complete design in hand. The combination of unique structures, and varying geologic and market conditions result in a lack of reliable and salient published historical cost information. The DB project delivery method presents even more unique challenges to the preparation of an engineers estimate and schedule, many of which are not encountered with traditional DBB projects. First, due to the nature of the DB approach, both the project delivery date and the proposal or bid date are advanced, reducing the time in which an accurate assessment of the project cost needs to be completed. Second, for most DBB projects, the engineers estimate is not finalized until project design is complete. Design-build projects require similar results from a design that is only completed to the 1030% level. Estimating uncertainty is relatively high at these low design levels, and even the most detailed cost estimates contain significant assumptions. For initial budgeting purposes, answering the cost question at a high managerial level is typically accomplished by comparing the proposed project to similar projects and then applying historical unit costs from those projects along with some cost index factor for escalation. However, this method is very limited in its usefulness beyond this stage because these rough order-of-magnitude (ROM) numbers generally have limited or no supporting information provided to validate them. Additionally, too many ROM estimates used for budget evaluations, particularly those derived from past costs, are based on total expenditures that include many unidentified risk items and contingencies particular to those jobs. Accordingly, the uncertainty associated with such historical costs is unknown. Finally, the use of historical cost data provides little if any information related to the scheduling of work, and the rate at which the work can be performed. The facility plan budget, like those of most projects, contained ROM estimates for each of the contract divisions. The PCOs first step in validating and refining the BPT estimate and schedule was to understand the basis for these estimates. The overall estimated cost for the BPT seemed reasonable, and portions of it were developed with a bottom-up approach, but because of its preliminary nature those portions did not go into a high level of detail. It quickly became apparent that a more detailed estimate needed to be assembled as a tool for evaluating alternative designs and the feasibility of reference designs. Our solution to this problem was to generate a bottom-up estimate with an integrated schedule. Building the Model Since the BPT is part of a much larger scheme, there are many interconnected elements of work and multiple contract units or divisions. Optimizing the interfaces between divisions was an ongoing process, with the BPTs alignment constantly shifting to best suit the needs of concurrent and follow-on contracts, and the demands of third-party stakeholders. Estimating work in this fast paced DB environment requires a

Design-Build Estimating from the Bottom Up 47

mathematical model of the project, where the effects of changing one input variable can ripple throughout the entire estimate. It would have been nearly impossible to keep up with the rapidly changing scope and configuration without such a model. In short, the operating descriptor for the estimate model in a DB environment is flexibility. Similar to the preparation of DBB bottom-up estimates, our first step was to break the project down into individual construction activities. Breaking the work down gave us an understanding of the required logic ties and the impact of individual design refinements. Performing quantity take-off on a conceptual design required us to rely on our experience in approximating finalized project quantities; the challenge here was to accurately anticipate and quantify what we could not yet definitively measure. We set up the take-offs so all known variables and assumptions were clearly noted and explained; thus, the variables were readily changeable as the design progressed. Establishing the estimate framework was perhaps the most daunting task and required the most flexibility, since elements were constantly being added, changed, or removed. Initially, there was no design, and many assumptions regarding means and methods had to be made. However, these assumptions were kept very clear. We took a moderately conservative construction approach for each of the project tasks to avoid building unnecessary negative bias into the base estimate costs. For instance, we assumed that the shafts would be initially supported using slurry diaphragm walls and excavated using a clamshell bucket. We also assumed an earth pressure balance tunnel boring machine (EPB TBM) would be used to excavate the tunnel. An earlier alignment shift had generated a lack of geologic information for portions of the tunnel; however, previous data had indicated to us that the use of EPB TBMssuch as those used in the construction of some Washington Metropolitan Area Transit Authority (WMATA) tunnelsto excavate the tunnel would be a reasonable assumption. Using these and other assumed means and methods, we performed cycle time analyses for each activity and built out the model with construction crews and equipment carried over from estimates of similar projects in order to expedite the process. With the selection of the design-build approach, the responsibility for the initial design of the BPT evolved into the development of a reference design, which was set forth as the constructible option, leaving the creativity to the design-builder. As it developed, this reference design became the basis of the estimate. We changed the assumed variables accordingly, and as design definition improved, the estimating uncertainty was reduced. However, waiting for this reference design to be completed was not an option. We used this time to update the estimate resources and independently check the quantity take-offs and estimate calculations. Using the Model The discussion makes clear that significant effort was put in to assembling a true bottom-up estimate, especially in the early phases of design, and the benefits of the modeling approach began to reveal themselves as the design advanced. As mentioned earlier, the BPT owes its existence to the addition of a program to reduce the nitrogen level in the Blue Plains effluent. To accomplish this goal, an enhanced nitrogen removal facility (ENRF) is to be constructed at the Blue Plains facility. Like the BPT, the ENRF has a deadline to meet for operation, and unfortunately, this means that it must be constructed at the same time as the tunnel. Accordingly, construction traffic became a big issue with several large projects underway in close proximity to each other. As a part of the bottom-up process for BPT, we had already estimated the number of segment deliveries, dump trucks, concrete deliveries, and so forth, so we were confident in predicting construction traffic for any given phase of the work. This saved the consultant performing the traffic coordination study significant time and effort, and was also beneficial in supporting the BPTs environmental documentation.

48

Contracting Practices and Costs

With the model, alternative evaluations were simplified. For example, while working with the dewatering pumping station designers to optimize the pump configuration and shaft size, the model aided in determining the impact on shaft costs for various pump layouts because adjusting variables such as shaft diameters was relatively simple. Other benefits came when dealing with third parties. The design-build approach necessitated that much of the interaction with third parties be accelerated, in large part due to the location of the BPT alignment. One of the concerns that rose among several of the stakeholders was the timing of tunnel excavation below their respective properties. By integrating a construction schedule into our detailed estimate, we were able to quickly generate the answers to such questions. Another advantage of having an integrated schedule in our model was that we could set realistic interim milestones for liquidated damages based on our current schedule. These milestones were tied to relevant activities. In this manner, the milestone dates were achievable based on our clearly stated assumptions with an allowance for float and would automatically be recalculated each time the design changed. Dealing with Uncertainty The cumulative effect of uncertainties in both cost items and schedule can have a major impact on the overall project cost. Items such as TBM production rates obviously have a major impact on a tunnel project, but are all too often entered in the historical record as overall average numbers (such as 15.2 m [50ft] per day), which can be very misleading to someone preparing a schedule or an estimate. If these historical rates include learning curves and time for scheduled maintenance stops, the actual production rates can be significantly higher. Since our objective was to be as realistic as possible when looking at productivity, we dealt with productivity uncertainties by performing a cycle time analysis. We evaluated the time required for each of the most basic production elements, such as segment erection and instantaneous penetration rates, and thereby developed a sustained average production rate. To gauge the accuracy of the model, we then compared our developed rate to daily production records from similar jobs, stripping out data outliers such as prolonged breakdowns and record advance periods. After establishing this daily rate, we added in delays for maintenance stops, learning curve, and shaft break-in and break-out delays. Using this approach for each activity, we and our reviewers knew what our schedule and estimate were based upon, which proved useful during subsequent risk management workshops. Risk Management As a part of the early design process, DC Water assembled an independent panel of industry experts to review the design approach taken by the PCO. This Project Review Board (PRB) was tasked with reviewing design assumptions and assessing the reasonableness of the cost estimate and schedule. A risk register was also developed for the project with input from DC Water staff, the PCO, and the PRB. Many individuals from each of these groups filled out a risk survey; these were then compiled and the items of the most significance were ranked in terms of their effects on both cost and schedule. We categorized these perceived risks within their respective risk model. Then, during a two-day risk workshop, all significant identified risk items were evaluated. Having a bottom-up estimate and schedule was very beneficial during these discussions because participants could see exactly where the dividing line existed between the potential additional risks they were being asked to evaluate, and the costs and durations already contained in the bottom-up estimate. These clearly defined assumptions prevented double-counting during the quantification of risk elements. In addition, the estimate and schedule provided guidance when it came to quantifying many of the uncertainties associated with the risk input variables, significantly improving the quantitative risk

Design-Build Estimating from the Bottom Up 49

assessments. The cost estimate and schedule thus served to help develop the @Risk model for project risk, which in turn was used to quantify reasonable cost and schedule contingencies for the project. These contingencies added to the base bottom-up estimate and schedule, and helped us validate the budget and schedule duration set aside for the BPT in DC Waters master cost loaded schedule. LESSONS LEARNED Pricing Information With the high levels of price volatility for materials and equipment in recent years, many local venders still seem reluctant to give pricing information or quotes to design firms because of the long time until actual construction, increasing the uncertainty in the cost estimates. Many of these venders do ultimately provide information, but generally only after multiple contact attempts requiring time not available for a DB estimate. It should be recognized that there can be significant uncertainty in the quotes received for major materials and specialty construction services; analyses of DBB bid results has revealed that the engineers number can be higher than the range of quotes given to bidders. Accordingly, the uncertainty that the engineer receives truly competitive quotes must be considered. Need for Coordination On a design-build project, good communication and early involvement between the design and estimating staff are crucial to avoiding potential redundancies and keeping the design and estimating processes running concurrently and efficiently. For the tunneling world in general, a bottom-up estimate with an integrated schedule can serve as a valuable tool in the design-build process, but in order for this value to be realized, the relationship between the designers and estimating staff needs to be one of understanding and communication. Many designers, especially those from outside the tunneling industry, do not understand that tunnel estimators are not like many surface structure estimators, who measure quantities and look up costs in a database. Good tunnel estimators have an understanding of tunneling means and methods, scheduling, and a vast array of other knowledge drawn from years of practical underground experience. Many project managers, typically nonestimators, underestimate how long it takes to prepare a bottom-up estimate for a tunnel. Therefore, while there may often be enough hours in the budget, there still may not be enough overall duration in the schedule allocated to develop an estimate as needed. One unfortunate result of this shortcoming for many projects is that there is usually redundant work performed, such as researching sites for muck disposal, finding temporary power supplies, providing input for environmental impact statements, and helping to develop realistic milestones for the project schedule. Benefits The main benefits of the detailed estimate model concept include the confidence it helps build by establishing a realistic and documentable basis for project costs, its ability to quantify uncertainty in basic production-related elements, many of which are based on estimating judgment, and its use as a powerful risk management tool. A detailed bottom-up estimate can also be very beneficial when it is incorporated early on in the design process, such as at the preliminary stages of a DB project. Using a numeric model of the project simplifies alternative evaluation, which helps to identify costly alternatives before too much effort is expended. Also, analyzing the project at a

50

Contracting Practices and Costs

high level of detail at an early stage can aid with identifying constructability issues long before they become a problem, allowing time for the development of better alternatives. BIBLIOGRAPHY Bizzarri, Ronald E., Ray, Carlton, and Edgerton, William E. 2010. DCWASAs project delivery approach for the Washington DC CSO Program, North American Tunneling 2010 Proceedings, 593600. Littleton, Colorado: Society for Mining, Metallurgy, and Exploration, Inc. Edgerton, William W. (Ed.). 2008. Recommended Contract Practices for Underground Construction. Littleton, Colorado: Society for Mining, Metallurgy, and Exploration, Inc. Stolz, John M. 2010. Overhead and uncertainty in cost estimates: A guide to their review, North American Tunneling 2010 Proceedings, 725729. Littleton, Colorado: Society for Mining, Metallurgy, and Exploration, Inc. Stolz, John M. 2010. Show me the money: The real savings in tunnel contract payment provisions, North American Tunneling 2010 Proceedings, 541547. Littleton, Colorado: Society for Mining, Metallurgy, and Exploration, Inc. Reilly, John. 2008. Probable cost estimating and risk management contracting methods, North American Tunneling 2008 Proceedings, 576584. Littleton, Colorado: Society for Mining, Metallurgy, and Exploration, Inc. Reilly, John. 2007. Alternative contracting methods, 2007 Rapid Excavation and Tunneling Conference Proceedings, 418429. Littleton, Colorado: Society for Mining, Metallurgy, and Exploration, Inc.

S-ar putea să vă placă și