Estimating Time of Construction

In a highly competitive industry tardiness can result in a partial loss of market to the swifter competitor who reaches the customer first with the product

by H. Carl Bauman, American Cyanamid Co.

^ ^ F ALL the elements of the fixed capital cost estimate, figuring time is probably most important. Every day which can be saved in completing the project assures early return on the capital idling during the con­struction phase. For the chemical industry turnover ratio is approxi­mately 1—a dollar of invested capi­tal usually yields a dollar of annual sales. A delay of a month in bringing a $1 million plant on stream means a loss of about $90,000 of sales.

Time enters all phases of the esti­mate. It becomes critical in the de­livery of key pieces of equipment and materials. It is the significant variable in the calculation of labor costs and field expenses. Estima­tion of time must consider the perils of weather, labor availability and productivity, strikes, transportation delays, design, and construction errors. There is available in the literature much good data on man-hours required to perform specific tasks. Most of it, however, repre­sents average conditions for wide geo­graphical areas and at best can be used as points of reference to be fitted to the particular project at hand. Local contractors can build up fairly accurate time records based on intimate knowledge of the labor market and special conditions in their areas of operation. Such in­formation represents valuable "know-how" of the contractor who is understandably reluctant to di­vulge his data.

Estimates of time of construction tend to be optimistic. A survey of a representative number of re­cently completed projects (Table I) showed that only 6% were completed ahead of schedule, 29% were on

schedule or no more than 1 month late, 37% were up to 6 months behind schedule, and 28% up to 12 months behind schedule. In­terestingly enough, there appears to be a correlation between dollar value and complexity of the project with delay in completion. Comple­tion of project is that time when the plant is first on stream producing 90% or more of capacity on a con­tinuous basis. Physical construc­tion prior to start-up operations is generally completed at or near the scheduled dale. Delays shown in Table I reflect largely the time re­quired to correct errors, omissions, and equipment and material failures encountered in starting up the chemical process plant. (See I /EC 52, 51A, March 1960.) The actual construction cost during this period is negligible compared with the cost of lost production.

In an attempt at determining real­istic completion dates for plant con­struction projects, fixed capital cost of a number of recent projects were plotted against completion time in months (shown at right). The re­sult is a pattern showing a correlation of project cost with time to complete. The study included a representative group of plants of the fluid, fluid-solid, solid, and civil-mechanical varieties. Capital costs varied from $0.1 million to $38 million. Plants costing less than $1 million took from 6 to 19 months to complete (Table II) . Median time was 11 months. The range was 1 3 to 27 months for plants costing $1 million to $5 million with a median time of 19 months. It took a median of 27 months to complete plants over $5 million in value. Time ranged be­tween 19 months for a single process

Project Average % Completion Capital Total

Time Value per Million S

Cases

Ahead of schedule 0 . 7 6

On schedule (no more than 1 mo. late) 2 . 0 29

Up to 6 mos. behind schedule 2 . 3 37

Up to 12 mos. behind schedule 11.0 28

Capi ta l Costs vs. Completion Time

Points shown here were deve loped from 35 major projects. Apparen t d ispar i ty in the lower range may be due to dif ference in t ype of construction contract

VOL. 52, NO. 8 · AUGUST 1960 4 7 A

I/EC Costs

Table I. Survey of Recently Completed Projects

COSTS CONSTRUCTION SCHEDULE

How is the construc-truction schedule do-i n g r This t ype of chart is in­va luable in keeping tabs on al l phases of the project. The chart is typ ical o f those kept by major companies

PROGRESS CHART

H o w d o e s t h e a c t u a l c o n s t r u c t i o n s c h e d u l e c o m p a r e w i t h t h e e s t i ­m a t e t i m e o f c o n s t r u c ­t i o n ? A chart such as the one shown here wil l help in deciding when and where a more deta i led analysis of the many phases of the 'construc­tion is needed to overcome time lags that have devel ­oped since the project was started

4 8 A INDUSTRIAL AND ENGINEERING CHEMISTRY

Table II. V· and

ariat ions Completi

in C a p i t a l ion Costs

Costs

Fixed Capital Value, Million $ C

Time in Months ompletion Median

1 and less 1 to 5 Over 5

6-19 13-27 19-42

11 19 27

fluid plant to 42 months for a com­plex chemical, fluid, and fluid-solid project.

As might be expected civil- and solid-type plants were found to go on stream earlier than fluid and complex chemical projects of equal capital value. Projects contracted out for construction on a cost plus basis were completed in two to eight months longer than those of equal cost let on a lump sum basis. The incentive advantage in the lump sum type of competitive construction bid seems evident.

Scheduling Construction

To assure that construction is completed in the shortest time possi­ble it is imperative that the work be carefully scheduled as soon after project authorization as possible. The project control estimate con­veys in essence a construction time schedule. Installation costs are based on the estimation of time in manhours to assemble the process plant. Other items of cost affected by time such as weather conditions, craft labor, travel, start-up and premium time are also incorporated in the estimate. A projected sched­ule of construction can be pre­pared from the control estimate by coordinating and synchronizing in­stallation work units in a sequence in which experience dictates they will be performed. Logically, site and substructure work will start ahead of building and equipment erection. As a consequence, de­tailed engineering and drafting must be concentrated in these areas ini­tially. Once "out of the ground," engineering work can be spread so that parallel designs proceed for all other phases of the work. Despite such planning, drawings invariably issue in a predictable order because of the dependence of certain designs on preceding work. For instance,

piping drawings cannot proceed until equipment layout has been sub­stantially finalized. Electrical work is dependent on equipment layout and to a large measure on piping drawing completion to avoid physical conflicts of location of piping and conduits, lighting fixtures and sprinkler heads, controls, and valves.

Flow of completed drawings and specifications proceeds generally as:

Site work Substructures Equipment and building super­

structures Piping Insulation Instrumentation Electrical

This flow of information establishes construction time starts for each class of work. Progress will be concur­rent in all areas as time elapses but starts will stagger depending on drawing release time. Having fixed the starting time for each category, actual duration of the patricular construction period is taken from the manhours allotted to that work in the control estimate. Replicas of typical estimate detail forms illus­trating manhour allocations are shown and discussed in I /EC 51, 68A (July 1959).

A typical construction schedule is shown on page 60 A. Estimated construction time for each work item is shown as a horizontal bar next to the work unit. Each bar represents a manhour total taken from the control estimate. A parallel bar differentiated in a suitable fashion from the estimated (hollow vs. solid, etc.) is employed to visualize periodi­cally actual progress. Each work item is assigned a weighted average value which is a ratio of its estimated manhours to total job manhours. Over-all project progress is shown as a single bar representing expected job duration.

Many graphic devices are em­ployed to make the construction schedule a valuable cost control aid. For instance, the chart (top, page 60 A) shows expected completion per­centages on the estimated (hollow) bar. The actual (solid) bar shows percentage completion only at the heavy vertical line which repre­sents the current report period. If progress is ahead of schedule the actual bar extends beyond the cur-

rent da te . I t falls short of the date line if the work item is behind schedule. T h e char t also shows by symbols indicated in the legend such information as the expected and actual dates of requisitions, purchase orders, and material delivery.

T h e detailed construction sched­ule is the basis for the preparat ion of a cumulat ive construction chart used to show job progress at a glance. Such a char t is illustrated on page 60 A and takes the shape of an ogee or S curve. Progress is shown by the dotted curves of actual completion compared with the solid lines repre­senting estimated performance. T h e horizontal distance between esti­mated and actual curves, to the t ime scale shown at the bot tom of the chart , reveals how far the job is behind or ahead of schedule at the report date .

As the construction schedule is always based on the job estimate, it should be revised only when the estimate is changed to reflect addi­tional work of any kind not originally contemplated. T h e schedule is not changed, however, if actual j ob progress proves the estimate to have been in error. A device used to help control costs, the schedule must be thought of as a means of pinpointing depar tures from expected progress so that adequate steps can be taken in t ime to keep within the forecast. Often a low estimate has been kept within bounds despite a slipping schedule by extraordinary but timely measures taken in re-engineering or reappraising construction tech­niques. A very safe schedule based on an overestimate is an invitation to overspend on the project. Early indications that progress is better than scheduled should be taken as a cue to re-evaluate the estimate and reduce the predicted cost, if justified, to avoid a complacent at t i tude which could result in fixed capital cost greater than need be.

Our authors like to hear from readers. If you have questions or comments, or both, send them via The Editor, l/EC, 1155 16th Street N.W., Washington 6, D.C. Letters will be forwarded and answered promptly.

VOL. 52, NO. 8 · AUGUST 1960 4 9 A

COSTS