Revision 3.1 - 49 - March 2010
transmitter is completely inoperative due to local failure (such as power loss, communication
loss, or complete electronics failure).
Commentary:
Transmitter and control valve positioners typically have a PID function block available, which
raises the issue of where the PID function should be located. Issues such as H1 segment
communications overhead, execution speed, advanced diagnostics, failure mode, and operator
access are generally considered when choosing where the PID block resides. However, the
general consensus is that the PID block should be located in the control valve positioner. As
with conventional control systems, loop and device failure modes need to be determined and
the proper course of fail action identified for each control loop.
If all function blocks of a single PID control loop cannot reside on the same segment, the PID
control algorithm shall be placed in the DCS controller. Cascade PID control can be
implemented in the field if all function blocks for both inner and outer loops reside on the same
segment. Complex control shall be fully implemented in the controller.
Commentary:
This restriction may not apply for systems supporting bridge capability between FOUNDATION
fieldbus H1 networks. Note that bridging may be possible with the use of HSE, but not all Host
systems with HSE support bridging. Note that there is limited user experience at this writing
and, therefore, the user should thoroughly test any implementation using bridging.
Each FOUNDATION fieldbus control strategy or module will be named as shown on the P&ID.
The primary loop function block used for operator interface (AI or PID) will share the module
name. With some systems, using consistent tagging conventions between devices and their
associated loop can facilitate improved access to device information.
4.3.1 Recommendations for Display and Alarm of Loop Variables for Control in the
Field:
Users will note that the measured/controlled value of any F
OUNDATION fieldbus PID loop can
potentially be accessed from a number of devices. When the PID resides in the final element
(valve positioner), the HMI can be configured to access all the loop variables from there.
However, a fault in that device will mean the measured value and alarms will cease updating
on the operator interface. In such cases, it is better if the host accesses the PV and alarms
from the transmitter (measuring device) directly wherever possible.
Commentary:
Users should also recognize that addressing “PV,” whether from the PID block itself or directly
from the AI (analog input FB), makes its update part of the “asynchronous” or non-deterministic
portion of loop communications (macrocycle), so display and trending latency may be
increased. The same is true when accessing “read back” (typically actual valve position as
seen by the positioner) from the loop’s AO block. Some vendors map the “read back” to the
AO block’s BKCAL_OUT parameter (“Use PV for BKCAL_OUT” checkbox) which then makes
it update with the deterministic portion of the macrocycle.