Can Topology Reduce Costs?

Nov. 3, 2008
It’s a Good Idea to Try New Topologies on a Small Project Before Embarking on a 5,000-Point Job
By John Rezabek

When plants install a network like Foundation fieldbus (FF) H1 or Profibus-PA, your consultant’s designers or your installation contractor might be inclined to default to a “tree” or “chicken-foot” network topology, where the home run or trunk cable ends in a grade-level, pipe-alley junction box (JB) and spurs to the field devices fan out from there. This is a comfortable strategy for both EPC designers and electrical installers, as the terminations in a given JB will fit nicely into a routine method of running conduit, pulling wires and landing terminations in the JB. Except for the trunk cable, the installation job from the JB to the field devices will bear a great resemblance to a traditional 4-20 mA or HART-based point-to-point wiring job. This is beneficial for the comfort level of all and could make the skill of the craftspeople doing the terminations slightly less crucial to success.

Process buses like FF and PA were designed to accommodate the existing networks and wires that one is likely to find in a legacy point-to-point 4-20 mA installation. But what if you have a new installation or you’re not aiming to reuse that old twisted-pair wire for the spurs?

I just had a project that used some Northwire single-pair Type A fieldbus cable. Since it’s made to a tighter spec, its O.D.—0.32 in. for 18 AWG—is pretty large for single-pair shielded. Once I get more than seven pairs in a single run, I’m already into 2 in. (53 mm) conduit. If you’ve ever done much estimating, you know larger conduit sizes not only cost more, but their fittings cost more, supports have to be stouter, and you get fewer in the rack.  It goes in much more slowly, with more man-hours/ft due to the weight and extra effort to bend it and hang it.

At the risk of driving your electrical craftspeople nuts, you could find some significant savings by exploiting the multi-drop network capabilities of networks like FF, PA and perhaps even Ethernet. What if every conduit leaving a field JB was no larger than ¾ in. (21 mm)? Reducing the terminations out of the pipe-rack JB will make it smaller, which will help to offset the cost of the additional junction boxes needed along the new extended trunk. Using a multi-drop method, the individual twisted-pair trunk winds its way among the instruments on its segment, with intermediate, smaller boxes and terminations as needed along the way. The investment in the design phase is a little higher and requires slightly more thinking both in the CAD phase and during installation. Whether this investment is offset by the dramatic reduction in the size and complexity of conduit runs depends on the size of the job. And there will likely be instances where a chicken foot remains the most sensible alternative.

Engineers would do well to make sure the installation contractor is onboard or that the scope of work is clear enough for bidders to consider that closer supervision might be needed. The electrical foremen will, for example, be less able to subdivide his crew into specialists who only run pipe, only pull wire or only stand in front of a junction box doing scores of terminations without moving around very much.

Ladder and basket tray can also make installation easier, as well as making future modifications much simpler. If you’re installing in a hazardous or classified area, you need to ensure that the tray installation satisfies the local codes. Some regions allow a high power trunk in a Div. 2 or Zone 2 area, but others might need that trunk to be intrinsically safe or otherwise current-limited. In the final analysis, compromises made to accommodate a less-protected trunk cable might outweigh the savings of a multi-drop topology.

Finally, even the much-maligned daisy-chain network can be used effectively when it makes sense, for example, connecting a network of indicate-only, multi-point thermocouple transmitters. But a daisy-chain should not be used where any control is happening, as no short-circuit protection can be used; the place where the daisy-chain connects to the trunk will potentially represent too high of a normal current draw for most short-circuit protection devices.

It’s a good idea to try new topologies on a small project before embarking on a 5,000-point job. We would aim to use our experienced homeys on the bigger jobs, in an advisory role if not as the key foremen.

John Rezabek is a process control specialist at ISP in Lima, Ohio.