In the ISA Automation Week Mentor Program, I am providing guidance for extremely talented individuals worldwide. We will be sharing a question and the answers each week. If you would like to provide additional answers, please send them to Susan Colwell at ISA. The first answer to the sixteenth question from Bahtiar Abu Bakar in Malaysia:
“What are the important criteria for selecting a system integrator to execute a DCS project for a Greenfield project (Cost, Experience, Timing)?”
Assumptions are the DCS has been selected, the process is well defined, the system integrator is only responsible for the DCS, and the process has both batch and continuous operations.
The first answer is from Hunter Vegas project manager at Avid Solutions, Inc. who I did a series of Control Talk columns with on Automation Projects.
In this particular case, you probably couldn’t go wrong in choosing the DCS vendor to do your integration. They’ll probably be fairly low cost and will probably do a good job given the relative simplicity of the project. Regardless of WHO you use, you need to be careful to specify a few items up front:
1. If you have certain ways that you want things programmed (certain dynamos, certain module templates, etc.), it is imperative that you VERY CLEARLY define that all up front. Don’t assume they’ll give you what you want. There is huge incentive for the company to use THEIR module templates and dynamos because that work is done and their labor costs will be very low (big profit). Customization costs them money, and they’ll try to avoid it.
2. Clearly outline the level of software documentation you expect and provide examples. Then ask for finished module examples early to make sure everyone is on the same page and you are getting the documentation you expect. (If you wait any length of time at all any change will require hundreds [maybe thousands] of updates rather than just a few.)
3. Interview the leader project engineer and his main engineering leads and determine exactly WHO will be on YOUR team. Be sure to lock THEM in on your project. Many integrators will “bait and switch” – showing you one guy and giving you another. The lead project engineer will likely make or break the project. Choose him wisely.
4. In a Greenfield job, there is usually a lot “going in circles” time where the project scope is very fluid and design decisions are still raging. If you bring an integrator in too early, they’ll burn a lot of hours working (and reworking) software trying to chase all of the changes. However if they are brought in too late then you may miss some good design opportunities or they’ll require a bunch of changes to get the DCS installed and functioning. I would suggest you hire the lead project engineer (only) to work on a part time consulting basis in the beginning to get the benefit of his expertise without killing a lot of money. Once you have finalized the design and gotten it moving THEN issue the purchase order for the configuration.
5. Demand that the configuration include basic simulation of the system and have the integrator provide a complete graphic (with all of the associated modules) as they are finished. Lease and/or buy a simulation for the plant and get the operators and engineers reviewing these early graphics very carefully as they are released. This is the time to make sure everything is right. (Tweaking one or two modules is easy. Correcting 200 modules that have been replicated off the bad ones is NOT.)
6. If equipment is duplicated then have the integrator fully program one unit and provide the modules and graphics for complete plant testing. Once this is done (and I mean REALLY done), THEN replicate the other units. Three copies of garbage is just more garbage.
7. If the job has batch involved make sure the team really does know batch and have successfully programmed it several times. They should have no problem providing samples of their code and a list of references. If they cannot, move on. You don’t want them learning on your job.
8. An integrator’s pricing should be competitive but I would definitely base my decision on a lot more than that. An integrator with a proven track record of successfully configuring projects like yours on time and on budget is worth WAY more than the relatively small amount of money you might save going with the “low cost guy”. A bumbled or delayed start up can cost millions of dollars and months of delay. A $50,000 saving will look pretty silly if you cost the plant an extra million dollars on the start up.
9. Integrator Size matters – to a point. If your job has any size to it then you probably want to avoid a very small integrator as they may be hard pressed to staff it. However, a very large integrator may not be good either – you’ll tend to have one guy who actually knows something and an army of fresh outs doing the grunt work. You want an integrator that has a proven track record of doing projects of your size so you can feel comfortable that they can handle the work and not go out of business tomorrow.
10. Look for an integrator that can add value…not one that just does what you tell them. A good integrator will ask lots of questions so they fully understand your process. If they see something that doesn’t make sense, they’ll ask why. They’ll volunteer ideas to improve process control, reduce maintenance costs, and save money on installation. They’ll also provide you with realistic schedules that they can meet. Be very wary of the guy who promises you everything…he’ll rarely deliver but won’t admit it until it is too late to do anything.