|
|
   
|
Prioritising Multiple ProjectsAP Van der Merwe Vice President (Programs) Project Management Institute (South African Chapter) Published in: ProjectPro, Vol. 5, No. 4, July 1995, p 44.
Abstract A qualitative method for gauging the criticality of a project, in order to determine the priority of that project against a range of similar projects, so that the correct measures may be applied to achieve the most efficient and effective use of available resources.
Introduction We all manage multi-projects every day of our lives as do many organisations, and we prioritise them without much ado. (11, 12, 15, 17, 18) So why the big fuss? My research showed that if the Pareto principle (17, pp 170-179) were true then the converse would prove that if projects were prioritised for effectiveness and efficiency 80% of them could be completed by using only 20% of the available resources. (26) Let me explain. Say you have 100 projects to manage. The Pareto principal states that 20% of these projects are going to consume 80% of your resources in terms of effectiveness and efficiency. We all know it to be true. Most of your projects are probably progressing well, but there are always a few which seem to attract every possible problem under the sun. These projects then use up resources to the point where the remaining projects will require additional resources. Alternatively, you find yourself in the position where after completing 30 projects you have consumed 70% of resources. (25) The Electrification Project is experiencing this problem, as are many other multi-projects. (10, 15, 16, 20, 21, 22) By not prioritising for effectiveness and efficiency you run the risk of tackling inefficient projects (in terms of their resource usage) early on in the program, resulting in the condition just described. (26) On the other hand, if you could prioritise projects for efficient use of resources into lets say three groups, then the inefficient projects could form one group, the most efficient projects another group, while the remaining projects could form the third group. Tackling the most efficient projects first means that some progress could be made and learning established before the inefficient projects are tackled. One could even debate whether the inefficient projects should be tackled at all.
Prioritising projects During my research (26) all the prioritising methods found (11-13, 20-26) were quantitative in nature in that some numeric aspect of the project was used and manipulated mathematically by a factor to form a priority. This usually took the form of project value or profit. Inevitably these projects were then sub-divided in geographical groups. I found the use of a factor resulted in statistically skewed results (18) in that the majority of projects would end up in one group. What is required is a prioritising mechanism that would allow one to vary the factor as well as enable one to change the parameter to which it is being applied. This cannot be mathematically achieved, therefore the tool must be developed in a qualitative manner. (26)
Prioritising matrix The problem that confronted me was the upgrading of 150 sub-stations on the ESKOM grid. (26) Eight parameters were selected to establish effective and efficient use of resources e.g.: Accessibility, geographical location, age of equipment, level of technology etc. Each parameter was given a range spanning three levels A, B and C as it was decided to prioritise into three groups as follows: 1. Accessibility A= Paved roads
2. Geographical locality A= <100 Km
3. Age of equipment A= <10 years
4. Level of technology A= Artisan
8. Importance of customers A= Interconnected at full load
(There is no limit to the number of parameters that can be used, but the range ascribed to each should be equal to the number of priority levels to aid decision making.) A matrix was then constructed as follows: Table 1
Detail withheld due to sensitivity of the information.
Priority for each substation was determined through group discussion (1, Modified Delphi) which applied a variable priority to various parameters. (Alpha characters were used to eliminate the urge to mathematically add the range for each parameter in order to obtain a total for each Substation.) For example: In the case of Substation 111 it would seem as if an overall priority should be C but in this case group discussion determined that Parameter 4 was of sufficient value to change the priority to B. In the case of Substation 221 the group decided that Parameter 5 was of less value but that parameters 1 to 4 were more representative of efficiency than 5 to 8 and rated this Substation B. In the case of Substation 357, group discussion determined an overall rating of A based on the effectiveness of Parameter 2. This resulted in a good statistical fit (18) by placing 20% of the Substations in priority A and 20% of the Substations in priority C. Further, all A projects represent the most efficient in terms of resource usage as they would be easily accessible, geographically close, new, have compatible technology etc. whereas C projects are inaccessible, geographically distant, old, have incompatible technology, etc. From this it can clearly be seen that the category C projects represent a nightmare of complexities, and in this case it was questioned whether these substations should not be replaced rather than upgraded. This resulted in the category C projects being withdrawn from the multi-project, reducing estimated expenditure as well as completion time by 60%. Applying the Pareto principle and its converse results in a 20% - 60% - 20% rule where the one set of 20% will contain solutions and the other set of 20% will contain problems.
Further use Selecting parameters to reveal a specific feature can easily be effected. The matrix was successfully used to determine the importance of 200 facilities (24) for maintenance purposes and has been demonstrated to determine efficiency for as many as three million projects. (25) Once the methodology is understood it can easily be applied to prioritise projects (or anything for that matter) with no limit to the number or levels of priority required. The most common range for prioritising seem to be within three levels, as it is easy to define these levels. One simply has to ask the questions: 'What is the most important, what is the least important and what would constitute the middle ground'. Once these questions have been answered it once again becomes relatively simple to determine the parameters which are considered important to these projects. In determining the range for each parameter one has to ask the same three questions as before 'What is the most important, what is the least important and what constitute the middle ground' All the required information is now available to construct the matrix, and when completed, ones colleagues/management are requested to help formulate the overall rating of each project. The results obtained in the manner described above have been applied to diverse areas within ESKOM achieving remarkable results. (24, 25, 26) In one of these examples an operating and maintenance group has used the prioritising method to perform an analysis based on substation importance and geographical location. The results of this analysis has impacted on the placement of maintenance personnel. This has led to a reduction in the response time to servicing of electrical equipment at the more important substations and therefore the outage time of faulty equipment, which has, in turn, impacted favourably on revenue earned by ESKOM, as reflected by the increase in plant availability figure in the annual ESKOM Report. Maintenance schedules are now drawn up to emphasise the priority rating of a substation, leading to more work being done to an improved standard on the more important substations. This has resulted in a dramatic reduction in the number of unplanned outages from 25 system minutes lost in 1990 to less than 5 system minutes lost per annum at present, again reflected in the annual report.
General learning principles The first principle learnt is that from the success achieved in prioritising 150 projects within a company, some improvement in the measure of effectiveness and efficiency can be expected from prioritising even as few as 10 projects. (25) The second principle is that to try and tackle 3 million projects without considering some kind of prioritising is a recipe for misfortune. It was Gen. George Patton who said that even a poor plan well executed is better than no plan at all. However we may have perhaps in a management context proven the Chinese general Sun Tzu (500 BC.) wrong, when he remarked "All men can see the tactics whereby I conquer, but none can see the strategy whereby victory evolves". The essence of project management is to be organised. However, to attack multiple projects without prioritisation epitomises disorganisation. Further, project management (contrary to popular belief) constitutes managing the people who manage the work and any project managers who tries to manage the work will soon find himself in difficulty. The third principle learnt is that if people are able to measure importance they generally tend to perform better at the more important tasks, (9, p371) resulting in a knock-on effect in efficiency. This effect can readily be seen by the improvements experienced in the application of prioritising as discussed above.
Bibliography
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
COLLABORATE, NEVER COMPROMISE |