Owner oriented information systems for multi-project control
Booz Allen & Hamilton
The advent of the 1970’s brought substantial changes to the field of owner project management. Contracting practices shifted away from fixed price contracts, increasing the burden of project control on owners. Investment programs trended towards larger projects, increasing the management problems of both financing and project control. Developing nations with limited management and technical resources became prominent as owners, while in developed countries, public sector involvement redirected and increased both information reporting and project control requirements. These and other changes created a new breed of projects with their demand for a new approach towards project management systems.
A large number of major investment programs currently being contemplated or executed are characterized by some or all of the following features:
- large capital investments
- integration of different production process and technologies within a single expansion program
- multiple contractors engaged to carry out design, procurement and construction activities within a single project or program
- diversity of contractual terms tailored to project and market conditions
- complex financing arrangements which may constrain contractor selection and/or impose stringent control requirements
Management of these large, complex, interrelated projects presents a new set of problems to owner management. Regardless of the amount of delegation to third party engineers and contractors, the ultimate responsibility for the creation of physical assets, on time and within budget, remains fully with the owner. In order to successfully discharge this responsibility the owner must efficiently manage, coordinate and control multiple sub-projects, in different phases of execution, delegated to multiple contractors under different contractual agreements. These tasks represent a new challenge to most owner organizations.
Prerequisites to Adequate Multi-Project Control
To successfully conduct a large expansion program the owner must successfully execute three key tasks.
1. Assignment of Responsibilities
The owner must first define how responsibility for the successful achievement of project activities will be assigned, both within his own organization and among contracting parties. This task includes the establishment of an approach towards type and diversity of contractual terms to be sought. A wide range of options are available to the owner approaching this problem, and a significant number of factors influence the outcome of the final decision. These factors include technological, schedule and cost constraints, market conditions as they affect contractor availability, and external government or financial restrictions. Assignment of responsibilities therefore must normally be tailored to the specific corporate and external environment as well as to the individual expansion program.
2. Establishment of an Expansion Organization
The second major task facing the multi-project owner is the creation (or adaptation) of an expansion organization geared to
- execute the project activities for which the owner has retained in-house responsibility (basic engineering, purchasing, expediting)
- monitor and control the cost, quality and schedule of activities assigned to contracted partners.
As in the case of assignment of responsibilities, the owner faces a wide choice of organizational options related both to the project characteristics and to the treatment of operational versus functional responsibilities. In a true multi-project environment, however, some form of matrix organization is normally best suited to owner requirements.
Figure 1. Multi-Project Matrix Organization
3. Implementation of Project Management Systems
Having divided responsibilities and created an organization the owners third managerial task consists of implementing project management systems. In a large expansion program these systems will exist at three levels: contractor, contract management, and overall expansion management. At each level procedures must be implemented to permit
- the smooth execution of routine activities
- the monitoring of cost and schedule information as a basis for control
- reporting of project information to the next higher level of the project hierarchy.
The three types of procedures encountered at each level of the expansion program are depicted in Figure 2.
Figure 2. The Three Levels of Project Management Systems
In effect they combine to form a three-tiered project reporting, monitoring and control pyramid as shown in Figure 3.
Figure 3. The Three Levels of Multi-Project Control.
The owners involvement at the three levels of reporting systems is necessarily different. At the contractor level he would ensure that his contractors possess adequate control systems, and that he has access to the information generated by these systems. Systems for contractor control are largely routine, and most owners have a well established set of operating procedures at this level, except where they face new contract arrangements, or are undertaking expansion for the first time.
The large multi-project expansion program, however, introduces a system requirement at the top management level which many owners are not equipped to handle: it is to this particular need that the remainder of this paper is devoted.
The Top Management Control Dilemma
Within the framework of multi-project management described above, the owner faces a complex monitoring and control task at the top management level. Cost and schedule information must be monitored on multiple projects or sub-projects in different states of completion, each employing one or more contractors under diverse contractual terms. Multi-project management in today’s world is accompanied by a high degree of informational complexity.
One such project with which the author is familiar was the construction of an integrated manufacturing facility and related infrastructure costing in excess of $3 billion. This project consisted of 34 individual facilities each having separate contracts for design and equipment supply, civil construction, and equipment erection. In addition dozens of contracts were let for the supply of support services in engineering assistance, expediting, transport, site facility operation, testing and construction management. Although six sub-project managers were appointed to coordinate large distinct segments of the project, top management still faced the problem of keeping track of the cost and schedule status of the enormous activity which the project generated.
Management of such a project must be predicated on an information system which is owner-oriented, flexible, timely, and simple. This system should
- structure information for owners and at a summary level of detail
- adapt to contractors internal project procedures and to a variety of contract terms
- report the best available current information
- operate with a minimum of staff
Contractors’ systems do not normally possess these characteristics. They are geared to managing at a greater level of detail than owners require, and are unnecessarily cumbersome, even slow, when applied to large multi-projects. More importantly, they are tied to the contractors’ internal operations and unsuitable for managing a multiplicity of major contracts.
Similarly, available software packages satisfy only a part of the owner’s requirements. Although they process whatever data the owner is able to obtain they fall short of solving the more general need for an integrated approach towards data gathering, structuring and processing.
An Integrated Approach
Through its project management consulting practice, Booz, Allen & Hamilton has been repeatedly faced with the owner control problems described above. In response to a clear need on the part of our clients, we have developed a unique approach to multi-project management information systems which has proven to be highly successful in meeting this need.
Viewing management information systems as links in the cycle of action/reporting/decision/correction, (Figure 4), such systems may be seen to perform the three tasks of data selection, structuring and processing. The success of a system is a direct function of its ability to perform these tasks efficiently and to orient output to the needs of the downstream decision-making process.
Figure 4. Multi-Project Management Control Cycle
A multi-project, such as the one described above, generates vast amounts of data. An efficient project management system prunes this data so that only essential information is processed for top management reporting.
Although this principle sounds simple, it is surprising how many project control groups develop and adhere to a “need to know”, precisely and in detail, “what is going on in the field.” On large projects, this tendency can be fatal. Only by reducing reporting requirements of sub-project managers and contractors to a minimum, can top management be assured of receiving timely and accurate data which can be processed efficiently.
An underlying concept of this approach is a strong decentralization of decision-making authority at the sub-project management level. The sub-project manager must be free to interface with and manage his contractor(s) as appropriate to the particular scope of work, contract type, and contractor systems. Project success relies to a great extent on flexibility at this level, and standard procedures should be designed to interfere as little as possible.
One useful technique adopted for simplifying reporting of cost information is the “Budget Amendment.” This is a manual procedure for reporting actual or anticipated cost changes as soon as they become apparent. All project staff are requested to fill out a simple form when potential significant cost changes are noticed, and to send it to the project control office. The success of this approach has been remarkable.
By correctly structuring his project code of accounts the owner can uniquely identify all project cost, schedule and resource elements using a single code. This is a significant asset to a multi-project management information system which
- simplifies both manual coding and data processing
- permits automated long range cash requirements forecasting based on integrated cost and schedule files
- facilitates extensive central analysis of unit prices and contractor productivity where quantity data can be obtained.
The coding approach should also meet the needs of corporate financial management, by identifying for each appropriate project cost item
- sources of fund
- asset/expense category, consistent with the corporate chart of accounts
- payment currency
Finally, the coding approach should facilitate contract control by identifying the contractor and work order associated with each cost item. A standardized work order procedure for sub-projects is an important feature of successful contract management on multi-contractor programs.
Automated data processing is a practical prerequisite to adequate owner control of a large multi-project. The volume of data generated, even when reduced to the minimum needed by top management, is simply too large to handle manually.
Several multi-project software packages are available on the market. However to maximize system flexibility a new software package was developed by Booz Allen & Hamilton having the capability to generate analysis based on fully integrated coding of cost and schedule. This package was also designed to interface smoothly with manual project control procedures such as the Budget Amendment and Work Order procedures described above.
An automated cost system should have the capability to store, update, and summarize information on:
- original, appropriated, committed and forecast budgets
- expenditures, advances and retainages against contracts
The cost system should operate in multiple currencies on international projects, and should retain sufficient historical data to permit analysis of changes to each cost element, as well as audit trails on individual expenditures.
Automated scheduling should be based on precedence diagramming. Although there are complex pros and cons for both precedence and i-j diagramming, precedence allows cost and schedule integration whereas i-j, for practical purposes, does not. This is an overriding consideration. Sophisticated scheduling systems allow multiple network end dates, negative lags and negative float.
Automated cash flow forecasting should give emphasis to three-month cash requirements. Long range forecasts should be adjusted automatically with budget changes and schedule updates, and should reflect the best information available to the user.
Owners of today’s multi-projects face a complex control task which cannot be adequately supported either by conventional information systems or by software alone. These owners require integrated management information systems which are owner-oriented, flexible, timely and simple. An efficient multi-project management system should tailor data selection, structuring and processing to meet these needs.