Scribd
Upload
58 views21 pages

Keeping Alarms in Check at US Oil and Refining

US Oil & Refining implemented an alarm management system called LogMate to help reduce the number of unnecessary alarms. They tracked key alarm metrics over time to identify problem areas. Their goals were to reduce steady state alarm rates to under 1 alarm per 10 minutes and upset alarm rates to under 20 alarms per 10 minutes based on industry benchmarks. Through iterative analysis of alarm data and working with operations and engineering teams, they aimed to eliminate unnecessary alarms tag by tag. An example problem of high frequency level alarms in an amine stripper was addressed through adding moving average logic to dampen the level signal in the control system.

Uploaded by

Sebastian Octaviano
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
58 views21 pages

Keeping Alarms in Check at US Oil and Refining

US Oil & Refining implemented an alarm management system called LogMate to help reduce the number of unnecessary alarms. They tracked key alarm metrics over time to identify problem areas. Their goals were to reduce steady state alarm rates to under 1 alarm per 10 minutes and upset alarm rates to under 20 alarms per 10 minutes based on industry benchmarks. Through iterative analysis of alarm data and working with operations and engineering teams, they aimed to eliminate unnecessary alarms tag by tag. An example problem of high frequency level alarms in an amine stripper was addressed through adding moving average logic to dampen the level signal in the control system.

Uploaded by

Sebastian Octaviano
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 21

KEEPING

ALARMS
IN CHECK
at
US OIL & REFINING

1
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

General outline of alarm management (AM) at USOR:


•Alarm problems were identified a couple of years ago
•Investigation of an AM solution/tool was requested and started (2005)
•LogMate was selected and subsequently implemented (Jan 2006)
•Collection of data began and shortly afterwards Alarm KB added in
with design/operating limits data (Apr 2006)
•Metrics (based on EEMUA 191) were identified and tracked
•Goals selected for acheiving benchmarks by analyzing data vs.
processes and working with various teams
•Tag by tag, alarm problems are being identified/reduced

2
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

USOR Factoids:
•Over 1800 DCS Tags (1450 are unsuppressed)
•2 Primary Units – Crude and Reformer (7-9 sub-areas each)
•2 ”A” Operators (one per unit)
•4 Alarm Priority Levels: Low, High, Critical, & Suppressed
•Each priority level has its own colour code and audible sound
•Critical alarms are displayed on separate, dedicated screens

3
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

Priority Levels:
Our priorities are allotted accordingly:
•Low = 80% of tags
•High = 16%
•Critical = 4%

These are consistent with EEMUA guidelines

4
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

Examples of our alarm counts in the Crude Unit per


month since inception of LogMate:
TOTAL ALARMS PER MONTH

AREA/UNIT FEB MAR APR MAY JUN JUL AUG SEP OCT

C-2 337 224 755 390 49 50 2 72 177

C-6 0 0 0 0 0 0 0 0 0

C-7 1205 188 544 774 1059 292 176 0 0

CLE 1116 339 210 2204 747 372 68 214 305

LCU 2419 1896 2800 1792 2403 1556 907 1645 1943

LCVU 2147 668 398 326 1831 333 751 2020 1693

PROC 7224 3315 4707 5486 6089 2603 1904 3951 4118

UTIL 540 358 234 362 248 198 370 318 280

CRUDE 7764 3673 4941 5848 6337 2801 2274 4269 4398

5
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

EEMUA 191 Long-term Steady-state Alarm Rate Benchmarks:


Per 10 minutes:
>10 alarms Very likely to be unacceptable
5 alarms Likely to be over-demanding (industry average)
2 alarms Manageable
<1 alarm Very likely to be acceptable

6
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

Corresponding Alarms per 10 Minutes:


ALARMS/10 MIN PER MONTH

AREA/UNIT FEB MAR APR MAY JUN JUL AUG SEP OCT AVG

C-2 0.1 0.1 0.2 0.1 0.0 0.0 0.0 0.0 0.0

C-6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

C-7 0.3 0.0 0.1 0.2 0.2 0.1 0.0 0.0 0.0

CLE 0.3 0.1 0.0 0.5 0.2 0.1 0.0 0.0 0.1

LCU 0.6 0.4 0.6 0.4 0.6 0.3 0.2 0.4 0.4

LCVU 0.5 0.1 0.1 0.1 0.4 0.1 0.2 0.5 0.4

PROC 1.8 0.7 1.1 1.2 1.4 0.6 0.4 0.9 0.9 1.0

UTIL 0.1 0.1 0.1 0.1 0.1 0.0 0.1 0.1 0.1

CRUDE 1.9 0.8 1.1 1.3 1.5 0.6 0.5 1.0 1.0 1.1

With an EEMUA benchmark of 1.0 for an ”acceptable” steady-state rate,


some months are good, some are over.
Obviously there is work to be done to improve our alarming.

7
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

EEMUA 191 Major Upset Alarm Rate Benchmarks:


Per 10 minutes:
>100 alarms Definitely excessive
20-100 alarms Hard to cope with
<10 alarms Should be manageable

USOR has set a goal to try and minimize upset alarm rates to <20
alarms/10 min

8
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

Our First Step in alarm management:


Begin the iterative process with reports to Operations/Engineering:
1. Generate weekly reports in LogMate to be sent out by e-mail
• Frequency Analysis
• Critical Priority Alarming
• Nuisance Alarms
• Relative Analysis

2. On a monthly basis, generate an Excel report of tag process performance versus


Design Limits
3. Evaluate periodically other reports or information to generate (e.g. Alarm Limits
versus Design Limits)
4. Maintain a ”log”/”score sheet” of evaluated loops for tracking AM progress

9
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

Our Second Step in alarm management:


•Evaluate the reports daily, weekly, and monthly in conjunction with
TRAC and Activity Analysis
•Pick out upsets or times of known events (e.g. instrument/unit work)
•Look for loops consistently alarming at high rates (we’ve selected 35
times/wk) over several weeks
•Record activities on and refer to ”Log/Score Sheet”

10
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

Looking at major upsets:

11
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

Looking for daily upsets:

12
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

Compare alarm frequencies week by week:

13
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

Compare Alarm Limits vs. Design Limits:

14
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

Our Third Step in alarm management:


•Work with Operations and Engineering to evaluate consistent ”bad
actors” and eliminate unnecessary alarming
•Look at process conditions and other factors
•Evaluate our design limits for ”bad actors” or unneeded alarms
•Make adjustments to alarm or design limits or to process conditions as
applicable

15
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

Our Fourth Step in alarm management:


•Evaluate on a regular basis our monthly alarm densities versus
benchmarks
•Look at the ”bigger picture” and re-focus/narrow in our attentions as
applicable

...After that, go back to the First Step and start iterations again

16
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

An Example of an Alarm Problem:


•The bottoms level on our Amine Stripper has high-frequency swings; as
a result, the high alarm is frequently activated
•The level controller for this controls a flow that bypasses the tower
bottoms but can cause resistance in the bottoms draining (backing up)
•The variable that directly affects the level is the reboiler (vertical heat
exchanger) and its configuration, causing level instability in the tower
bottoms
•Currently we cannot physically/mechanically change anything, so a
logic/programmable solution must be sought

17
18
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

Our Current Solution to the Alarm Problem:


•Switching level control functions with another valve having direct effect
on the bottoms level was not accepted...
•Changing the high alarm setting is neither practical nor effective...
•And adjusting alarm deadbands in our DCS is not a viable option...
•...SO... a Moving Average block was added to the LC input logic to
”dampen” the level signal, thus mitigating the ”false alarm” situation (until
something else could be done)

19
20
KEEPING ALARMS IN CHECK AT
US OIL & REFINING

From all we’ve done to date...


...What have we accomplished so far?
•Went from NO idea what state our alarming was really in to having a
gauge and means to measure our performance
•Defined a new Alarm Priority Level (”critical”) for our most important
alarming, with defined criteria to ensure these alarms are meaningful
•Evaluated all loops to set ”limits” on alarm levels (i.e. Design Limits) so
our alarms have some justification for their settings
•Made ourselves more aware of conditions of upsets or instrumentation
problems so alarms we get can make sense
•Opened up new communication between departments to help operate
smoother

21

You might also like