How system integrators, plant owners should collaborate on an automation project
Here are two pitches for the same control system integration project:
This new process unit will be a technological showcase: no control room-everything will work from hand-held human-machine interfaces (HMIs); software in the cloud and maximum connectivity via the web. Wireless sensor networks will be everywhere.
Or perhaps something different:
This unit must be reliable, so only tried-and-true technologies will be used-basic analog field devices, all hard-wired; standard PC-based HMIs like the older systems at this site. We know what works, and we will stick with it.
Obviously those two approaches represent extremes of the technology spectrum. The question is, who is making the suggestion in each case? Is it the plant owner who is always conservative while the automation system integrator wants to push the envelope? Or vice versa? The answer can go either way, but in most cases the situation lands somewhere in the middle. There might be some separation on the spectrum, but it’s not on opposite ends.
Separation, to any extent, has to be worked out, and the technology gap has to be closed. Sometimes it’s an action requiring patience and compromise on both sides, but when done well, it will improve an automation project.
Isn’t the customer always right?
Some project discussions begin with the owner taking a strong position and telling the integrator: "I am the customer, and it’s your job to do the project the way I want. Here is the scope, fill it as described, and charge appropriately."
This approach works for projects where the scope is tightly defined, and there are few options for forming an alternative solution. Many automation projects, however, become more complex for a variety of reasons. Entering into a more open discussion goes a long way toward a successful conclusion.
The table lists reasons for involving an integrator in the technology selection and implementation process, and these points are discussed in detail below.
Choosing the DCS
The number of process plants launching major control system upgrades, including full distributed control system (DCS) replacements, is increasing. Many companies trying to run plants using 30-year-old systems find the risks and costs of nursing such old platforms intolerable. Problems caused by breakdowns and replacement part shortages interrupt production, and vulnerabilities to outside cyber attackers pose major threats.
When a DCS must be replaced, the first hurdle is supplier selection. For some companies, the choice is simple: for reasons of history, supplier relationships, corporate engineering policy, or whatever, there is only one choice. The DCS will come from Supplier X. If this is the case, any system integrator working on the project will have to live with this decision or not accept the project.
For the owner, stating the selection at the outset might thin out the number of integrators bidding on the project. An integrator not familiar with Supplier X has to make a choice: Bow out or take the project believing it can train internal personnel quickly or hire appropriate talent. In most cases, the integrator will not be able to pass those extra costs to the client and must look at it as an investment. Hopefully, another project will come along where those new skills can be used again.
In a situation like this, a large integrator with extensive experience across a range of suppliers and platforms has a distinct advantage. It’s simply a matter of assigning the right people to the project.
If supplier selection is open to discussion, an integrator can play a critical role. The process begins with the plant owner’s engineers examining operational needs: Here is what the system must be able to do as a minimum, and here is what we would like it to do in addition to improve process control.
The integrator’s project team looks at these requirements, and internal platform experts representing each DCS supplier draw on knowledge and experience to make a case for why a particular system best meets the owner’s requirements. When a consensus is reached, the best one or two possibilities are presented to the owner by the integrator with a clear explanation as to why specific recommendations are being made. Often the integrator’s cumulative internal experience plays a major role in the process. Memories of where a given platform worked well or poorly in particular applications are very important and guide the selection as a result.
Having this unbiased experience is paramount for such a decision. The internal engineers at a given facility don’t have the time to research products along these lines and in necessary detail. If they have only the sales pitches from DCS suppliers to go on, their view might not be entirely accurate. Because the integrator often has hands-on experience with the platform in a variety of applications, it is easier for the integrator to achieve a higher level of objectivity than the owner can alone.
If the integrator feels the customer’s selection is not the best for the situation, the internal expert for the specific platform will explain the reason for the misgivings. The integrator will point out where gaps exist between functionality needs and the system’s capabilities. If the owner accepts the evaluation, but sticks to the original selection, the next step is to figure out how to overcome the deficiencies. The integrator’s role in this process is critical.
In some cases, the integrator’s project leader may have to sit down with the DCS supplier’s representative and be very clear as to where problems lie, allowing the company an opportunity to offer a solution. If the DCS supplier sees the plant owner as a critical customer, it may be glad to have the opportunity to solve the problem and further cement the relationship. The alternative may be for the integrator to bridge the gap with its own fix, either independently or in concert with the DCS supplier.
Learn more about challenging DCS situations and what to do when opinions clash.
The toughest DCS situation
The most difficult situation for everyone is when an entire plant changes hands. The new owner with its engineering staff moves in to find an environment where it had no prior involvement whatsoever. That is, no control of supplier selection, no influence on control strategy, and existing intellectual property is a black box. Some of the plant’s existing engineers may have remained behind and can help, but it’s hard to depend on this kind of cooperation over the long term.
An integrator working with a new owner is in the same position of bewilderment at the outset. The task is to make sense of what is there, identifying which systems need help and how the automation infrastructure can be improved. This is another situation when having an integrator with a deep technological bench is important. The particular plant may be unfamiliar, but if there are experts within the staff capable of working with the major process units and automation systems, moving through the learning curve can be much faster.
The new owner and integrator need to analyze the entire plant, from one end to the other and evaluate the processes and supporting automation systems. As problems emerge, they need to be prioritized according to their ability to disrupt production. Those bad actors need to be fixed first to stabilize the operation.
The next step is to look at where improvements can be launched to improve throughput and gain efficiencies. These are normally treated as incremental programs, marching methodically through the systems one-by-one while the plant remains in production. It’s rare to see the new owners of a working facility shut it down and make all the improvements at once.
The same guidelines for technology selection also apply. The owner may decide to leave the existing systems in place for cost reasons, or it might want to bring the new plant up to its current engineering standards. Changes will be based on a variety of influences while balancing costs and benefits. The ability to analyze and quantify those steps is where an integrator can be very helpful, bringing an external view to the situation combined with practical experience with other similar plants, processes, and automation systems.
When DCS opinions clash
Most integrators with any amount of experience can remember situations where a client wanted something contrary to the advice from the integrator. An engineer for the owner may be convinced of an approach, but the integrator disagrees.
Sometimes the owner’s approach does not reflect the integrator’s first choice, but it still represents good engineering practice and should be sustainable over the long term. In such a case, the integrator will probably state its opinion but will proceed and implement the solution.
Other situations may be more serious because the owner’s choice could be untenable in the integrator’s view with respect to plant performance, safety, a critical operations parameter, or nonconformance to accepted industry best practices. In these scenarios, the integrator has to be clear and honest in explaining the reasons why that approach doesn’t work as well as the likely outcome for the long term. If the integrator views the situation as something capable of causing an incident or other unsafe situation, it may be necessary to refuse the project and document the reasons why.
These kinds of cases are rare, but they do happen. Few plant operators become so stubborn as to insist on a truly destructive course. Generally as things move up and down the chain of command more moderate views and sensible plans emerge. An integrator pushed to follow a valid approach different than the preferred choice should look at such situations as learning opportunities.
Long-term DCS collaboration
A more common situation is where an integrator is asked to follow a strategy appropriate for the situation but does not believe the plant owner can sustain it over the long term. It’s unfortunate, but in this post great-recession environment, and with oil and commodity prices at low levels, producers have cut costs to the bone and are often operating with minimal plant operations staffs.
When an integrator finishes a project and hands it over to the customer, the moment of transfer might be the best the plant runs. In spite of the operator and personnel training, when the integrator is gone, the owner may not have the wherewithal to sustain the improvements made.
A few years ago, this was a common occurrence. Most companies had not realized the extent to which all the personnel cuts were hobbling their effectiveness. The truth has set in over time, and producers realize they need help. They still operate with minimal staffing, but they understand the need for getting help from external resources; this often means engaging the integrator’s services over a longer term.
Many automation projects no longer have a defined beginning, middle, and end—so an ongoing relationship is needed to maintain peak performance and ensure safe operation. The integrator stays on the site in one way or another to ensure gains made with the project are maintained as well as to suggest and implement improvements. Individual operators get to work with the engineer who designed the solution, and the integrator can optimize and fine-tune the project over time.
Plant owners who understand the value of production in dollars and cents realize what an effect such an ongoing relationship can have. The ability to increase throughput or avoid an interruption can have a huge positive effect, far beyond the cost of this type of service arrangement with the integrator.
When an integrator knows it will be on a site for the long haul, it has a different stake in the project and technology selections. Both sides—integrator and owner—are now working together with the knowledge that finishing a project is the first step.
This has an influence for selecting an integrator as well. If developing this type of relationship is critical, the integrator has to have the capability to support it. There have to be enough skilled people available to have one or more available for the client at all times. Just as a deep bench is necessary for evaluating and applying the widest variety of automation technologies, it is also necessary when providing personnel resources over the long term. The ideal situation is where key individuals responsible for designing a solution can remain in place to support it.
The most effective projects and relationships are those where the plant owner and system integrator cooperate as trusted partners, instead of just as a vendor and a customer. This partnering relationship creates the highest value for both sides.
Tim Gellner, senior consultant, Maverick Technologies. Maverick Technologies is a CFE Media content partner. Edited by Chris Vavra, CFE Media, Control Engineering, firstname.lastname@example.org.
Many automation projects become more complex for a variety of reasons; entering into an open discussion goes a long way toward a successful conclusion.
When disagreements occur, the system integrator needs to decide whether there is room for compromise and to decide whether the changes being asked are acceptable.
The most effective projects and relationships are those where the plant owner and system integrator cooperate as trusted partners.
What other aspects of a project should a system integrator consider before deciding to embark on a project?
See additional stories about system integration projects below.