Hydrocarbon Processing - April 2021 - 82

Environment and Safety
TABLE 1. Criterion for the acceptability of alarm systems
No alarms recorded in stable operation

Acceptability Section A to D: no additional source

More than 1/min

Unacceptable

1 every 2 min

Over scathing

1 every 5 min

Manageable

< 1 every 10 min

Acceptable

FIG. 2. Getting data considerations.

the level of acceptability are established,
which will be the main criterion for the
measurement and subsequent control of
the alarm management to be implemented. Many ways exist to establish this criteria; however, having a reference based
on recognized standards or good practices will provide greater support and a
future possibility of reference with other
similar processes or plants.
As an appropriate and well-accepted
criterion in the industry, the Engineering Equipment and Materials Users Association (EEMUA) 191 standard1 can
be adopted, which sets the level of acceptability depending on the number of
alarms recorded under normal conditions (TABLE 1).
Depending on the level of acceptability determined during the diagnosis, the
goal is to improve the level as implementation progresses, recommending the
" manageable level " to be achieved (i.e.,
one alarm every 5 min).
Control systems usually have functionalities that allow the alarm records to
be obtained, which, in turn, are properly
organized according to the process variables (flow, pressure, temperature, etc.)
to which specific process or equipment
are associated, to which signal or controller they belong, and even the type
and level of alarms or automatic stop
(FIG. 2). This makes it easier to manage
the information for the analysis stage, al82

APRIL 2021 | HydrocarbonProcessing.com

lowing personnel to export that information to computer tools such as spreadsheets or databases. To avoid distortions
or inaccurate decisions, it is important
to consider the following when selecting
evaluation periods:
* Information should be taken
during periods when the process
has developed within normal
operation (i.e., without scheduled
or unscheduled plant shutdowns)
* Preferably under relevant operating
conditions (i.e., processing loads
or relevant production levels).
An important part of the diagnosis is
the validation of staff competencies that
carry out the process control (e.g., operators, panelists and supervisors) to identify the training needs and action plans.
Experience indicates that the fundamental documents previously listed are often
set aside, so they must be incorporated
with the most up-to-date versions for
new personnel training.
Analysis. This stage is based on the information obtained in the diagnostics
stage. Several types of deviations can be
found in this stage, which depends on
the particularities of each case (process
type, management style, organizational
culture, etc.). However, at this stage, it is
advisable to group the variances into the
following aspects:
1. Rationalization. It is important
to be clear that rationalization is
not an alarm removal process but
a process of managing alarms in
the right place at the right time.
Depending on the configuration
of the control systems, there are
processes that are controlled
in a certain place and monitored
in the main control room. For
example, a truck loading station
usually has a special control
workstation-different from
the main control room-that
is only monitored. However,
redundancy is usually incurred

when setting alarms, causing them
to appear on both stations. This
condition can lead to a lack of
operational discipline, since being
at both stations, the operators
can expect the other to attend to
them and vice versa, or worse, the
alarm is not attended after a long
period of time. For this type of
case, it is advisable to establish
a level of criticality of alarms that
should go in redundancy towards
the main control room, given its
level of impact on the process
and its security or its
interrelationship with the other
processes of an installation
(processing, storage, etc.).
Other types of rationalization
examples can be found in the
alerts of the same system (warning
of upcoming maintenance,
capacity alerts, etc.), which are
usually seen as alarms and distract
the attention of operations
staff. Usually, the architectures
of the control systems have
maintenance or support modules
managed by the maintenance or
instrumentation personnel. These
types of alarms can be managed
through these modules only,
allowing them to clear the control
screens of the process itself.
2. Review of inconsistencies
and failures. It is important to
identify the nature of the alarms
that are presented on a recurring
basis during normal operating
conditions. The main causes to
be analyzed include:
* Changes in process
capabilities (processing load
increases, changes in operating
conditions, different modes
of operation, etc.)
* Inadequate calibration of
equipment or instruments
* Uncontrolled changes in alarm
settings in the control system
* Errors in the process design.
This review requires a detailed
analysis in the engineering
and design of the equipment,
especially in the instruments and
equipment associated with the
recurrence of alarms. For example,
it is common for an automatic
cutting or emergency valve to set

http://www.HydrocarbonProcessing.com

Hydrocarbon Processing - April 2021

Table of Contents for the Digital Edition of Hydrocarbon Processing - April 2021