Parallel play: An old and new control system share information and saveWritten by Jeremy Pollard Monday, 03 March 2008 00:00
Thirty days of downtime and $21 million in lost product value. That's what Motiontronix, a Wonderware Certified Systems Integrator, was able to save its customer.
This story tells how they realized these incredible results by using OPC tunneling, bridging and a server to allow an old and new control system to operate in parallel and share information so operators would always know what was happening during a changeover.
Motiontronix’s customer is a large food processing mill in a remote location that operates 24/7. When the customer needed to upgrade its control systems, it also needed to minimize downtime and maximize production time. Motiontronix knew the best way to do that was to change the control system PLC hardware by running the old and new systems in parallel.
In previous installations, Motiontronix did the work in four days, working 24 hours a day, since the old and new systems could not communicate in parallel. The extra costs and risk of error from such long shifts were high. The biggest problem during the changeover period occurred when an operator selected an operation from the old SCADA and operators and engineers testing the new system could not see the changes. In an ideal situation, the team could check the new system before taking the old system offline and the systems could operate in parallel. But prior to this installation, it hadn’t been done before. Could Motiontronix make it work this time?
According to Rudi Van Der Merwe of Motiontronix, "the overall system normally consists of about seven to eight PLCs with approximately 3,500 I/O points. Assuming you just take one minute on each I/O point for checkout, you need about 60 hours to change from the old to the new system if there weren’t any problems. The reality is there are always problems, so changeovers would take even longer. Production time lost in the conventional method would have been a minimum of one to two months."
How did Motiontronix do it? By using OPC tunneling, bridging and a server provided by Software Toolbox. "The software is quick and does not slow the overall system down, and you do not have to rewrite tags," Van Der Merwe says. "This makes it possible to change each PLC as we like and keep the systems running in parallel. The other big advantage is that if you encounter any problems with the new system, you could switch back to the old system within 15 minutes, nearly like a total redundant system. We lost no production time, versus a typical 30 day production loss in a typical changeover. We saved our client $21 million on a small mill."
The solution uses a TOP Server, OPC Server and OPC DataHub tunneling and aggregation software to minimize downtime and configuration time while increasing performance. To complete the upgrade, Motiontronix used the TOP Server Mitsubishi Suite driver with its SuiteLink and OPC interfaces to provide data to the new InTouch application and the OPC DataHub to allow the two SCADAs to share data from the single driver across the network, without using DCOM, a key consideration for reliability and ease of configuration and setup.
This was accomplished by using one OPC DataHub node as a DDE Client on the PC with the existing system’s Wonderware Modbus I/O server. The second OPC DataHub node was used on the TOP Server PC as an OPC client talking to the TOP Server OPC and I/O Server. Using the tunneling, aggregation and bridging capabilities of the OPC DataHub, the team was able to mirror the data across the network without DCOM and bridge between the OPC and DDE servers, thus sharing data between the SCADAs and allowing both systems to run concurrently.
"With us now getting the two SCADAs sharing real-time information and the programming tested we can changeover within an hour. This is done when the customer stops production for their normal eight hour weekly maintenance per PLC," Van Der Merwe says. "Certain PLCs in the system are mission-critical for the operation of all the mills and the customer cannot afford to stop for long periods of time. There is also less pressure on engineering to complete each change over as all the wiring to get the two systems operating in parallel is done beforehand."