My automation-related blogs RSS feed alerted to me to a Control Engineering Asia article, Simplifying Connections, written by Emerson’s Nathan Pettus. I’d not yet seen this article published a few months ago.
In it, Nathan describes the challenges many project engineers face as they try to keep their automation projects on time and on budget. Late changes in I/O requirements can be a big problem in holding to the schedule. He describes some typical scenarios such as skids showing up on site with different I/O counts or different mixes of analog and discrete I/O than what was planned. One example is a skid containing analog level transmitters instead of digital level switches.
Nathan highlights some of the project repercussions:
…there is a likelihood that the spares in place are not of the correct type of I/O – e.g. maybe the skid has extra 14 digital outputs, but there are only 8 available in the I/O subsystem for the skid. Even if there is room for additional I/O capacity from a card perspective, projects are under tremendous pressure to reduce costs up front, and therefore there may not be any room in the I/O cabinet itself. And after working through all these complications, the drawings and wiring must be reworked, which depending on skill sets and available resources, may or may not be possible, but will certainly be expensive in both time and costs.
Typically, the project electricians bring the wiring into a common location from all the wiring junction boxes throughout the process. These wires are landed in a marshalling panel. From this point, they are crossed wired to the I/O of the control system or PLC, to the appropriate I/O card type (analog in, analog out, digital in, digital out, etc.) It can be hard to keep track of where all these wires should be landing and even more difficult if late I/O changes cause this cross wiring to be rewired.
Nathan describes the implications:
…if I/O that has already been cross-marshalled needs to be moved to another controller, the multi-core cables from the field have to be moved from one marshalling panel to another. Then the cross-marshalling has to be redone, which causes costly delays. This may also increase the need for panel space, and will definitely increase engineering time and result in changes to engineering drawings. Sometimes, it will be cheaper to just send the cabinet back to the factory to have it re-worked versus completing the work in the field.
Nathan contrasts the process with DeltaV electronic marshalling:
…the key difference is that there are no cross-marshalling wires; all of the design, and engineering work associated with crossmarshalling is eliminated, because each individual I/O channel will be electronically marshalled to any controller in the system. This means that the I/O can be bound to specific controllers at any stage in the project. If late changes are made to I/O types, or new I/O is added, no adjustments to existing wiring or the cabinets must be made. Instead, new I/O can be added to the marshalling cabinets and electronically marshalled to the controllers that need the I/O.
Also in the article, he shares other project implications such as cabinet space and junction box requirements, and how electronically marshalled I/O fits with digital bus-based I/O and wireless I/O.