IO module replacement advice: Plan ahead on instrumentation, space, design
Manufacturers and production facilities will often schedule an I/O upgrade to take place during their slowest time of year (for example, during the Christmas holidays) to keep any disruptions in production to a minimum. Yet for many businesses, such as water and power utilities, there is no “slow time.” Their operations are running 24/7, so they need to find a way to complete their upgrade as quickly as possible, while still keeping all machinery and equipment operating. The key considerations for accomplishing this are many, but smooth I/O upgrades almost always require a huge commitment in terms of proper planning, which should take into account things like interdepartmental collaboration, costs, time, personnel, and of course, wiring.
Proper planning is where everything begins. Too often, I/O upgrade projects are reactionary. Perhaps there are new and specific production problems, or maybe the control system is very old and, due to attrition, there’s no one left who understands it; or maybe the person or persons who designed the control programs have left the company, modifications need to made, and now there’s no one left who understands how to make the necessary software changes.
It’s not prudent to jump in and begin changing such major control components like I/O just because you’ve had to correct one specific problem or because the control system is now unfamiliar to current personnel. The business’s most essential operations are at stake, and there’s considerable groundwork that needs to be done first.
Part of the plan should be inclusion of key personnel from all the relevant departments within the enterprise. Production managers and engineers will naturally be involved, but an I/O upgrade needs to be properly budgeted for, which means accounting and purchasing departments also need to be consulted so all anticipated expenses can be covered (and allowances can be made for the additional purchases and expenses that often creep up). Networking and cabling will certainly be an issue, which will likely require the participation of the IT department to configure firewalls, allocate IP address ranges, and bring routers and switches online. Are we talking about I/O that’s controlling facility power, lighting, or water? This requires the input of company electricians, plumbers, and other building maintenance personnel as the upgrade is planned.
Good planning also includes contingency preparation. Is there a plan in place to handle any alarms that might be triggered when the old I/O connections start getting disturbed? Also, very often the variables and actuations for one specific I/O point will directly affect others. Cascading PID loop control, where the output of one PID loop is used as the input for another, is a good example. Simply upgrading one I/O point handling PID will have serious effects on production processes. Things like this have to be prepared for in advance.
Regarding planning, above all, it’s important to have the upgrade directed by a team who has a comprehensive understanding of the old and the new systems—from a hardware and software perspective. Without this base of knowledge to draw upon, replicating and/or improving the control programs and processes with the new I/O system will become more difficult.
Wiring the I/O: Details
Instrumentation: It’s important to consider the many often overlooked implications of altering I/O. Study the control system carefully and consider the outputs the I/O is handling and all the inputs that control is based on. In other words, it’s easy to think about how wonderful it will be to gain all new levels of control once the new I/O is in place, but often that control is contingent upon a number of analog inputs aggregated by various types of instrumentation deployed all across the facility. This instrumentation has to be taken into consideration before any type of upgrade can begin. For example, is the instrumentation compatible with the new I/O modules? Can the necessary connections be made? Simple things like this often get overlooked. For example, the middle of your upgrade is a terrible time to discover that the field signals or wiring for RTDs or thermocouples (in the industrial boilers or ovens, for example) come with special connectors or terminations that won’t connect with the new I/O modules.
Also, the new I/O may have different current or voltage range input specifications (0-5 V vs. 0-10 V, for example), which could also cause wiring problems and unexpected delays while safe and proper connections are established from the field equipment to the I/O.
Space: Cabinet and rack space is another issue. Can the cabinet or rack physically accommodate the new I/O and the I/O processors and controllers? Also, if operations will be continuing during the upgrade, will there be space to deploy this new hardware in parallel until the changeover is made? Furthermore, in scenarios such as these, sometimes the old I/O is so old that it’s become fragile. Wiring connections can be so worn or delicate that disturbing them can cause breaks or other failures. This needs to be taken into account before shifting or moving that old I/O to make room for the new racks.
Power: A quality I/O system requires quality power and good regulation of that power. (A very large share of calls coming into our product support line is for power-related problems.) Having an adequate power supply correctly and securely wired to and supplying the new I/O can help avoid faulty signal readings and communications-related problems that could hamper the I/O upgrade project.
Design: In general, control engineers are reluctant to initiate an I/O upgrade project and may be even more reluctant when they were not the ones who designed and/or implemented the current system. One reason is because they’re being forced to work within the constraints of a wiring scheme that they themselves had no hand in developing. When building a control system, there are many ways to design an architecture that will accomplish the desired tasks—and for each of those design architectures, there are many different ways to do the wiring. So when upgrading I/O, it’s important to study the wiring scheme beforehand to determine how well it will accommodate the new hardware. Maybe the old system was designed in a way where there are a lot of fixed I/O positions, where a specific location, a specific I/O rack, or specific slots on the I/O rack have been designated for all analog points, or all digital points. With the new hardware, will the same types of I/O be placed in these same locations? If not, then be prepared to spend extra time and effort rewiring.
One way around this is to choose new I/O that allows for any module in any position on the I/O rack. I/O of this type will also pay dividends when it comes to system scalability. As the application and the new control system expand (and they very likely will, as those working with them familiarize themselves with and discover their capabilities), having a more versatile, “any position” I/O module means it won’t always be necessary to permanently define ahead of time what functions (input or output, analog or digital) each and every rack of I/O is going to be handling.
Also with regard to wiring designs, it can make a considerable difference—and either increase or decrease the workload required to complete the upgrade—whether the field equipment and instrumentation is wired to a terminal strip or directly to the module. Again, it comes down to versatility, and having I/O that offers the option of direct connections (from the top of the module) to a wide range of sensors and devices can, in some cases, significantly shorten wiring runs.
Starting an I/O project? Ask your vendor!
Before tackling the foreboding task of sourcing and implementing all that new hardware and investing the considerable time and labor needed to rewire everything, go back to your automation hardware vendor and ask if it offers migration paths to upgrade its legacy I/O and control systems. With something as simple as a processor swap or network coupler replacement, the project gains all features and functionality of a full-scale replacement from all the old I/O: features such as Ethernet connectivity, multi-protocol support, and the ability to migrate the current installation to faster networks for higher performance and newer and better communications options.
For example, Opto 22 offers a migration path from its older, single-channel, serial-based I/O systems to its more powerful, Ethernet-based Snap PAC system. The company has a line of drop-in replacement boards for old systems that upgrade them to Ethernet networking while adding support for multiple simultaneous protocols, including popular ones like EtherNet/IP and Modbus/TCP. Instead of a large-scale I/O replacement, control engineers and field techs can simply remove the older processor or network coupler from the I/O rack, replace it with the new one, and the system will run continue to run. No modifications to existing I/O racks, modules, instrumentation, or field wiring are required. And as a bonus, after the upgrade, the company’s newer, more powerful software can be implemented, giving users more sophisticated control and data acquisition capabilities like PID control, subroutines, and OPC connectivity.
– Ben Orchard is application engineer with Opto 22. Edited by Mark T. Hoske, CFE Media, Control Engineering, www.controleng.com.