Skip Navigation LinksHome > Discussion Papers > 5 BIM in Procurement
Skip Navigation Links
HomeExpand Home
 

Transforming Construction?

5.0 Conventional Contract Procurement

Price competition is the driving force of the market economy. It ensures that buyers get the goods and services they require at the best possible combination of price and quality. In order to survive in this competition, suppliers are forced to innovate continuously, either to supply at the lowest price, or to supply the highest quality. Suppliers who are unable to compete are eliminated from the market; only the best survive. Thus, effective competition has two hugely beneficial effects for the economy: goods and services supplied at the lowest prices; and continuous improvement of products and services by suppliers

The lump sum, fixed price competitive tender is, in theory, the closest approach that exists to pure price competition in the selection of construction contractors. All other things being equal, this should be the best mechanism for arriving at the lowest cost outcome on individual contracts. It should also be the most reliable way to embed effective competition - and thereby real improvement - in the industry at large. As will become clear, competition in construction - intense thought it may appear to be - achieves neither of these goals.

The difficulty is that, for price competition to work effectively, the buyer must be able to specify his requirements fully, accurately and unambiguously, and must subsequently be able to compare potential suppliers' competing offerings on an exactly like-for- like basis. The documentation used in construction procurement does not enable this to happen.

The most systematic and in theory, the most precise form of procurement documentation used in construction is the bill of quantities (BQ). (There are many other forms of procurement documentation, generally less rigorous than the BQ, which all suffer even more than the BQ approach from the problems outlined here.) Although it may be used for other purposes in the course of the project, the central function of the BQ is to provide an accurate specification of the scope of work of the contract at hand, delineated in such a way as to enable accurate, like for like comparisons of the bidders' proposals.

In the standard BQ preparation process the taker-off locates, classifies and measures each of the individual building components, as he observes them on the relevant drawings. He then groups the details of these individual items into class aggregates, according to the rules of the relevant Standard Method of Measurement (SMM). For example if the project contains say, 120 reinforced concrete columns of varying cross-sections and heights, at different locations throughout the building, all 120 might well be described in three clauses: one for the formwork, one for the reinforcement and one for the concrete - "Concrete, class A, in columns - 928m3" for example.

Such drastic compression of the original detail is necessary because paper is the traditional medium of communication of BQs. Without this level of compression the project might have to deal with the storage, copying and distribution of thousands of pages of bills - a huge and cumbersome job. The computational power and storage capacity of modern computing systems means that none of this is any longer actually necessary, but the tradition persists. The result is that although the surveyor may record a great deal of potentially useful information about each of the individual components of the building, he must then effectively throw away almost all of that material in order to produce the sort of terse, cryptic, SMM compliant line items of which the example above is typical.

The fundamental problem with the BQ approach is that the compression process that's used in the creation of the bill is not reversible. It is not possible to map line items in the bill unambiguously to elements or components of the building as shown in the drawings. So, for a given bill item, even if one adheres strictly to the rules of the SMM, it's impossible to re-generate the component-level detail that gave rise to it. It is therefore impossible to ascertain the true scope of a particular line item, which in turn makes it impossible to know the true scope of the contract as a whole.

The contractors bidding for a project know that once they sign the contract, they are signing up to the scope as supposedly specified in the bill. In order to ensure that he has covered the scope fully, each of the bidders for a particular contract must therefore carry out his own take off and attempt to reconcile this with the buyer's bill. Given today's industry of layered contracts, sub-contracts, sub-sub-contracts and so on, the resultant duplication of tendering effort involved in procuring even the simplest building is astonishingly wasteful.

A second problem with the BQ approach is that, for the reasons outlined above, even the most diligently prepared bill contains a large proportion of line items that cannot be verified precisely. Each of the bidding firms interprets the bill in its own way and responds in its own way based on this interpretation. This makes it impossible for the buyer to arrive at definitive, accurately comparable responses. Instead, the procuring party, with each bidder, must carry out a variety of 'normalisation' or 'bid conditioning' exercises - adding a little scope and cost here, deducting a little there - massaging the numbers until he's got a proposal that he and the bidder can agree on. This will hardly ever be the real, lowest possible bid for the work, but no-one will ever be able to discover that, either to prove or to disprove it.

A third problem with loosely or ambiguously defined work scopes, as represented in bills of quantities, is that all the bidding contractors must factor in scope risk in their tenders. Regardless of whether any of the bidders is behaving in a predatory manner - win at all costs, then profit from claims - each of the contractors, if he wants the work, must assume that someone on the bid list will bid unreasonably low, either inadvertently or deliberately, therefore he must do the same. He cuts his core unit rates to the minimum, cuts his allowance for overheads and preliminaries, and cuts his margin. Everyone must do this. Incompetent bidders, who fail to see what's involved, bid low, get the work and lose money. Predatory firms who see the claims opportunities, bid low, get the work and make their profit from claims. Competent, non-predatory firms, who make reasonable allowances in their bids, bid higher and lose the job or bid lower and lose money. That, in a nutshell, is how competition works in construction.

The bidders end up competing for the claims opportunities, because there is simply no profit to be made in the core scope of work. Increasingly, competition in the industry is not between the construction competence of rival firms, but between the capabilities their respective estimating and claims departments.

The huge waste of effort and the generalised failure of conventional procurement techniques to achieve the lowest price from project to project arguably has little impact beyond the projects in question. However, the failure of the process to ensure the selection of the 'best' bidder is a far more pernicious problem for the industry as a whole. Best bidder in this context is taken to mean the firm most capable of performing the construction operations required. Less competent firms, with better estimating or claims capabilities survive, as predators, in this system. These are not the firms who should survive, or who would survive in an efficient market. But this is not an efficient market, functioning as markets should. To repeat: it leads neither to lowest price projects, nor to best of breed contractors. The result instead is a profitless, subsistence industry, with no capacity for investment in either physical or human capital.

5.1 BIM-based Contract Procurement

A construction industry in which contract procurement was based on Building Information Modelling would be an entirely different proposition. In a BIM model there exists a digital equivalent for every one of the components that make up the physical building. The model takes the form of a powerful database which can hold a great deal of data about each of these components. The data held includes information about the component's classification, as well as the details of its physical properties, its geometry and its location in the building. The database can be queried in a wide variety of ways. So, for example, it would be quite simple to extract a complete, detailed schedule of components, grouped by classification or trade and by location. A copy of this schedule, in database or spreadsheet format, together with the relevant drawings - or even better, the relevant model or model section - can then be sent to the bidding contractors, for them to price and return.

This is essentially how e-procurement works today, with the key difference that the component schedule provides a definitive, verifiable statement of the scope of work in terms of components to be installed into the building, by the relevant trade. Each of these components can be located, identified and itemised by the bidders, so their tenders will be exactly comparable at the component level of detail. There is no scope for gaming the process, and given that the design stage, as described above, delivers a complete design, there is no opportunity deliberately to bid low in pursuit of claims. Contractors compete on the basis of their ability to perform the work most efficiently, that is a lowest cost, rather than on their ability to chase claims.

In this scenario, the two key objectives of a competitive market are achieved: lowest price and best performer. And because the contractors don't have to worry about predatory bidders, they can pitch a reasonable price, including a reasonable provision for overheads and profit.

Ideally the model should be created during the design phase of the project, as a deliverable from the design programme, and used as outlined in Section 4 to support design communication. However, even if a BIM model is not produced during design, it is a relatively easy and very cost effective exercise to develop such a model for Procurement and Construction purposes. Once created, a BIM model provides a unified, coherent representation of the building that everyone involved in the project, and everyone with an ongoing interest in the operation of the building, can use and benefit from. One of the most valuable applications of the model is in planning and progress assessment on the ongoing project. This will be addressed in the following section.

 

GSK House