About eight years ago, ExxonMobil made a commitment to using main automation contractors (MACs) on large projects and those that involved multiple engineering, procurement and construction (EPC) firms. Timm Madden, senior consultant at ExxonMobil Development, joined Ron Day, project director at Fluor, and Roger Reid-Bicknell, project manager for Emerson Process Management, at the 2008 Emerson Global Users Exchange Engineering Forum to discuss the advantages and disadvantages of using a MAC approach on large projects, as well as the communication and scheduling disagreements that can arise.
The MAC concept is to identify all automation activities, develop an execution plan, obtain the best resources, select the best technologies and then let the resources do what they do best, said Ken Valentine, department manager at Fluor, who moderated the discussion. Madden represented the owner perspective in the discussion, while Day gave the EPC take.
Who does what is critical. ExxonMobils Timm Madden stressed the importance of well-documented scope definition and management under the MAC methodology.
Madden warned, however, that although demand is currently great for engineers, thats not necessarily good for MACs and EPCs because engineers often find a better arrangement elsewhere mid-project. Maintaining a qualified workforce is difficult, he said.
Madden, Day and Reid-Bicknell all agreed that data coordination is another potential hazard to the project. More and more, were compressing our schedules, said Madden, so the I&C data that is developed by the EPC and needed by the MAC is frequently late. In the ExxonMobil development plan, we include all deliverables to and from the MAC in the EPC and the MAC schedules. These two schedules need to dovetail together.
Other potential problemsknock-on effects of late facility design changes and the impracticality of integrated FAT because of multiple EPCswere identified by Madden. Most large projects have many interfaces, and many of those are with the MAC, said Madden. The I&C chain of command isnt always clear, and packaged vendor support and I&C technicians often are not available. Sometimes EPC process design engineering isnt even available.
From the owners perspectivein this case, ExxonMobilshow willingly have EPCs accepted the MAC concept? Many EPCs have figured out that MAC concept is a good thing for them, claimed Madden. Initially, there were a lot of complaints. But the hours that the MAC is taking away are the high-risk hours, and the work isnt among the EPCs core competencies. The MAC concept is still in evolution. Eventually, this will be the standard way of operating.
Both Day and Reid-Bicknell agreed with Madden that integration of the EPC and MAC schedules was important to project success. But each had a slightly different take on some of the other critical factors. Where Reid-Bicknell encouraged direct contracting between the MAC and the client and early definition and scheduling of the MAC role in the commissioning and startup, Day emphasized the EPCs need for a change management plan and the impact of EPC and MAC disbursed execution.
A detailed responsibility matrix for the owner, EPC and MAC needs to be developed, said Day. The more detailed, the better. We prefer an execution strategy that promotes the EPC having direct control of its own destiny. You need to ensure the selection process emphasizes evaluation of the MACs proposed execution plan, execution organization and infrastructure. For some reason, one of the most difficult things to do is to fully integrate the MAC and EPC schedules.
Day agrees with Madden on the need for coordinated schedules. We understand the tug of war over hardware and software data and deliverable freeze dates, said Day. The MAC wants that data earlier, but as an EPC, we often have difficulty meeting those dates. Any changes after the freeze dates create a commercial implication. Its a good idea to just plan for change. Ive never been on a project where we didnt have change.
In the end, however, owner/PMC oversight is required to audit consistency across multiple EPCs and to intervene should differences arise between a MAC and an EPC, said Day.
The MAC model has become the model for large projects or for projects with multiple EPCs, asserted Reid-Bicknell. Emerson has executed more than 50 MAC projects, and Reid-Bicknell explained there are four main typesthose contracted by a third party; those contracted by the client/EPC; those contracted by the EPC; and direct to client, the preferred type.
The MAC needs a clear identification of deliverables critical in scope definition, he said.
The benefits of the MAC outweigh its disadvantages. Extensive interface agreements must be established when dealing with multiple locations, and there are no contractual obligations for the EPC and MAC to work together, so it takes some extra effort to create that kind of teamwork, explained Reid-Bicknell. But using a MAC eliminates some of the bid packages, and it ensures that the automation system is consistent with the clients vision as defined in the FEED [front-end engineering and design], and that the automation vendor is substantially more responsive to the clients need and vision. Plus, the MAC can offload some of the workload and provide leadership with interface management plans.