We're adding two new CNG pressure/flow regulating stations that will supply a new source of gas to supplement the existing CNG reg stations that feed four CTs in a 1,000 Mw CC plant. The new pipelines' regulating stations will be blended downstream of the existing pipelines' regulating stations.

Each regulating station will be controlled by its own INFI90 PCU. The flow monitoring will be done in a 3rd PCU. This is because that's the way the existing reg stations are set controlled. The I/O includes two regulation valves, two slam shuts, flows and pressures for a total I/O of 2 AOs, 2 DOs, 11 AIs, and 4 DIs (or possibly more I/O we haven't thought of yet).

The two existing reg stations are already controlled via analog & hard wired relays and there's plenty of room in the existing conduit runs. Wireless is out because it's not fast enough and this application is too critical.

To me, the advantages of the old hard wired approach are that it has a proven track record, it's easy for us to engineer in house because we're very familiar with INFI90, and we have enough spare I/O so we avoid having to add the hardware and drivers for three protocol converters.

The main drawbacks I see to the old approach are the labor and material costs to install thirteen or more miles of cable, and we'd be missing an opportunity to extract all the information that Hart, Fieldbus or X-bus? devices offer today. It would be nice to modernize, if for no other reason than to become familiar with new approaches.

If we did go with digital, the new positioners and transmitters could be bus compatible, but the old flow meters have only analog output. And what about limit switches, D/O relay control for the on-off slam-shut valves, and position feedback for the self regulating valves. I suppose there's X-bus devices that do all these functions, but I'm treading on new ground here.

So I'm looking for ideas and opinions on the traditional vs. the new age approach.

 

Flow meters (ultrasonic, vortex, mag, and DP etc.) also use Foundation fieldbus.

Most limit switches are eliminated since the fieldbus valve positioners already contain continuous position feedback and end-point alarm could be set if need be.

Pressure, temperature, level, and flow switches are falling out of fashion because they have no diagnostics. Use low cost fieldbus devices instead.

Position transmitter also come with Foundation fieldbus, although I have only seen one (which has not been FF registered yet): http://www.smar.com/products/tp302.asp

On-off "slam-shut" discrete valve couplers and actuators are available with Foundation fieldbus http://www.fieldq.com http://www.topworx.com

Some control valve positioners can be configured to operate in an on-off mode http://www.fisher.com

Analog weaknesses:
. Closed loop inaccuracy
. No integrity check
. Limited power
. Range mismatch
. Range limited
. No innovative devices
. No firmware download
. No alerts

In my personal opinion, this is why:
. Closed loop inaccuracy
Transmitters are digital, control systems are digital, and positioners are digital. Does it make sense to have analog signals between them? D/A in transmitter, A/D in input card, D/A in output card, and A/D in positioner. In each conversion a little bit of accuracy is lost to quantization error and calibration mismatch, etc. This adds up for the loops of the plant units. And the plant units add up for an effect of overall plant output. Digital transmission eliminates this analog problem because value is transmitted undistorted.

. No integrity check
If analog signal changes from 19 mA to 18 mA due to ground loop or too high wire resistance this cannot be detected because 18 mA is a valid signal. Likewise you cannot tell if signal changes is process or noise. Digital transmission eliminates this analog problem because value has status and error checking.

. Limited power
There is less than 4 mA for the device to operate. This limits the level of sophistication for the devices. You can find diagnostics and features in fieldbus devices not found in analog devices.

. Range mismatch
If the range of a replacement transmitter is set incorrect the reading, control, and alarms in the system will be incorrect. For example if range is 0-400 in system but is accidentally left as 0-800 in the transmitter, the display in the system will be half so alarms and control will be incorrect. Digital transmits the true engineering unit value and is not distorted by range mismatch.

. Range limited
For an analog transmitter the 4-20 mA range is usually set much narrower than the minimum and maximum sensor limits. For example, a transmitter may be capable of measuring -15 to +15 but range may be set to the normal operating 0-7 for best accuracy. If during abnormal condition the level shoots up to 13 then for any value above 7 the output is 20 mA (or 20.5 mA) so system display will be clamped at 7 (or 7.2). Digital transmits the true engineering unit value within full sensor limits and is not clamped within a narrower range. Knowing the true level at an abnormal situation could be useful.

. No innovative devices
An analog transmitter only has one output and therefore can only do one measurement. Digital does not have this limitation. Therefore you can find 8 channel temperature transmitters for fieldbus. U channel temperature transmitters also exist. Similarly, pump health transmitters provide more information than can be conveyed through a single 4-20 mA value and are only available as fieldbus. Another innovative fieldbus device is a multi-point digital indicator.

. No firmware download
Analog transmitters do not accept firmware download so they cannot be upgraded to take advantage of new features. Fieldbus can.

. No alerts
Analog devices provide no failure alerts. This makes it hard or impossible to distinguish between process problems and device failures. Fieldbus makes this distinction clear. Operators can take appropriate action. Problems are pin-pointed and can be fixed sooner.

Cheers,

Jonas