The Riskenomics dependency modeling tool was used to illustrate the business risks to a major financial institution on a large City office construction project nearing completion. The project program was divided into two distinct phases, the base-build phase which included the installation of central plant, equipment and systems, followed by a fit-out phase.
THE CHALLENGE
According to the construction program, the base build phase was 3 months from completion. Both phases of work had independent project teams. The client had specified that the Data Centre and IT Main Equipment Room (MER) were to be operational by the end of the base-build work. Full occupancy of the building would follow thereafter on a phased basis, in line with fit-out progress. The building would ultimately house the clients Senior Management and core business groups including over 1,000 trading positions.
It was quickly established that the base-build phase of the project would not be completed within the remaining 3 months of the program. Any delay would have significant consequential business impact for the use of the Data Centre and MER and ultimately put the fit-out and phased occupancy programs at risk. The compiled dependency Completion Risk Model focused on the needs of the client and in doing so identified the following obstacles to project completion:
Earlier in the project engineering systems design compromises had been made in order to stay within the project budget, these compromises were identified to have increased future business operational risk to an unacceptable level.
The base-build contract had little chance of being completed on time, as project managers were relying on information provided by the various package contractors, whereas they should have made their own assessment of progress status and actions required to remain on schedule.
- The base-build project team was overloaded with project problems and was trying to identify what was important in an intuitive way rather than through a systematic approach.
- The fit-out team had relaxed because they believed the building would not be handed to them on time, therefore their work program was put at risk.
- There was no coordinated completion strategy.
- There was no coherent and integrated testing strategy to prove that all the building support systems worked properly and in a fully integrated manner.
- There was no validation strategy at the end of each project phase.
Although most of the information needed to complete the dependency Completion Risk Model was available from the project teams, there was no simple process or method in place to communicate project status and issues to the client.
OUR SOLUTION
The Completion Risk Model enabled the project team to focus on what had to be completed to minimize risk to the clients business. The model identified all outstanding work and unresolved issues that were significant risks and the necessary project team resource was provided to resolve or mitigate the risk to an acceptable level, in order that the base-build phase would be ‘business safe’ at hand-over.
The design compromise was a decision by the project team to omit a UPS module in order to keep within budget. To help mitigate the reduction in UPS capacity a `load shedding` control system had been designed that would limit the electrical load to essential services in the event of available power supply failure. The Risk Modeling process, being also a dependency model, was able to demonstrate that if this control system failed, the power to hundreds of dealing desks would be at risk.
DEPENDENCY RISK MODELING WITH RISKENOMICS
The dependency risk model identified what was important to the clients business. Evaluation of the issues to be addressed and the resources required to make progress in these areas, made it possible to agree an amended base- build completion program, which would provide an acceptable `safe building` status in which the clients Data Centre and MER could operate.
The Riskenomics dependency risk model was used to review the proposed fit-out contract and again showed that completion of this phase was also at risk. The fit-out project team was asked to compile a completion commissioning logic diagram, without allocating time to the commissioning logic activities. Once the logic diagram was established, time was assigned to each task and it was possible to demonstrate that completion could not be achieved in an acceptable timeframe using the current program of activities.
Consequently, the commissioning logic was changed to provide a more efficient program, one which would still achieve the essential commissioning tasks, but within the clients required completion program. The revised commissioning logic diagram and associated program listed each task and when completion was required, this in turn would drive the completion dates for when specific parts of the installation were required to enable the commissioning to be carried out, as the smallest of delays would put completion at risk.
Risk models were also used to identify building defects that constituted a risk to the client business operation, these would need to be resolved ahead of building occupation. Ignoring minor defects that required remediation, of which there were many, would seriously impact business operations. Risk modeling of important defects gave direction to the project team and enabled adequate resources to be applied, thus avoiding delay to the completion and future occupation program.
The dependency model showed a validation process had not been incorporated in the initial commissioning program. The Riskenomics models were able to demonstrate that even if all the constituent parts of what was a very complex project had been tested, the business would still be at risk unless full systems integration testing was undertaken. In this particular case a ‘black building’ test, requiring the main supply to the building to be isolated at source, allowing all the various back-up and safety systems to operate as designed. All would require test and validation.
The logic for performing full integration tests was accepted by the client, although the project team, who, due to a limited and shortened program, were under pressure to complete and handover, remained skeptical. The test was carried out, but only after a detailed method statement was prepared and approved, following full risk assessment and analysis, as by this time the data centre was operating and processing business sensitive data.
The integration test was conducted and within twenty minutes of isolating the main building supply, the standby generators failed and the full building electrical load, including the live data centre, switched to the limited operating capacity of the UPS batteries. The electrical mains were quickly restored and a subsequent post-mortem identified problems within the load shedding and standby generator control systems.
A decision had to be made on whether or not to postpone the imminent relocation of a large trading group to the new facility. Using the dependency risk model, it was possible to demonstrate that by using manual load shedding, the critical business operation would be safe until the automatic systems could be investigated and the operational problems solved. The client immediately provided financial approval for the installation of the missing UPS module and the control system problems were resolved. A further test proved the manual load shedding operation and the trading group were able to relocate, thus avoiding delay to the clients lengthy and complex relocation program.
The risk model approach had proved successful for the client and this methodology was adopted for both the management of risk and closure of the remaining outstanding works and project defects.
