|
BACKGROUND
Compressor
fouling and deterioration causes production
losses on many major platforms and disrupts
strategic supply sources for gas around the UK. Yet many other installations operate very
successfully without suffering fouling. Until
now, expensive refurbishments have
provided the only realistic way of countering
fouling. However, this frequently perpetuates
problems and recurrent production losses are
experienced in subsequent operation.
Getting
on top of compressor performance is central to
production. Yet what goes on inside
centrifugal compressors in the field is a
mystery to most, including the manufacturers.
So
how can a company be sure its equipment is
working optimally? And how does that
performance compare with others in the same
industry? For a field producing $500,000 a
day, a 15% loss in performance matters.
This
thinking led to MSE setting up
the first ever Joint Industry Project (JIP) for
optimum centrifugal compressor performance in
the hydrocarbon gas industry.
The JIP Pointing to the Optimum Centrifugal
Compressor for Hydrocarbon Gas Service was
launched by MSE in the UK to:
-
Decipher
the factors that characterise fouling in
compressors derived by the evidence
provided by analysis of the combined
experience of a wide population of
compressors used in hydrocarbon gas
compression service
-
Evaluate
the performance losses experienced by
centrifugal compressors, which so often lead
to loss of production and equipment failure
-
Choose
examples of compressors operated on major
production platforms where performance loss
in production is experienced, and quantify
the efficiency losses
-
Choose
examples of compressors with good
performance retention in production
applications and quantify their performance
-
Make
an analysis of the population of
compressors; isolate the effects of process
conditions, operational usage
characteristics and process system designs
-
Isolate
the effects of seal wear, balance piston
wear, diffuser fouling, blade fouling and
the internal recycle
-
Analyse
differences in centrifugal compressor design
criteria and isolate connections to fouling
effects
The
outcome would be a blueprint for the optimum
centrifugal compressor that was resistant to
fouling
and with maximum performance retention.
JIP:
PHASE 1
This
was the what. It involved collecting
operating data from some 50 compressors and
constructing a database to enable the machines
to be benchmarked and conduct an initial
performance analysis. MSE quantified head
and efficiency losses and produced initial
population trends. We found:
-
Up
to 25% head loss and efficiency loss
(compared with manufacturers’ data)
-
Power
discrepancies between absorption and
delivery
-
Certain
compressor designs may be more susceptible
to performance loss than others
JIP:
PHASE 2
Second
phase was the why. We showed:
-
The
effects of four key performance loss
mechanisms
-
Sensible
design decisions lead to performance
retention
-
Even
a low level of liquid carry-over leads to
sustained performance loss
Now
each JIP member can view online, via a secure
password, the performance of their own machines
and how they compare in detail with various
trends exhibited by anonymous data from the
other members of the JIP.
JIP:
PHASE 3
The
third Phase linked performance loss mechanisms
to compressor design attributes and produced a
series
of invaluable design guidelines, which have been
validated
against extensive field data collected during
the project,
for operators. It was concluded with
a workshop
and seminar for compressors users with the
theme:
' Regain Performance to Recover Production ’.
This stand-alone seminar was also open to non-JIP
members.
TECHNOLOGY
Under
this Joint Industry Project MSE has developed:
-
12
design guidelines for optimum design of
centrifugal compressors
-
The
database of over 50 centrifugal compressor
designs and real performance characteristics
-
The
Loss Mechanism Test - for diagnosing
causes of compressor degradation from
performance measurements
-
The
Health Index - for calculating the
risk of exposure to performance losses of a
new design
Design
Guidelines for Optimum Centrifugal Compressors
By
analysis of the design and operational
characteristics of over 50 compression trains,
twelve centrifugal compressor design guidelines
have been derived. The guidelines
direct engineers and designers in correct
machine selection. This ensures peak performance and
robust design to protect against performance
losses due to service-related problems.
Industry-wide Machine
Database
An
industry-wide Machine Database (currently over
50 machines) supported on the internet is now
accessed by the five oil companies to evaluate
their designs with the accumulated experience of
the growing population. Designers may
determine the optimum impeller geometry, speeds,
and impeller diameters. The database
identifies machine geometry for peak efficiency,
highlighting the machines that have the
greatest fall in efficiency with use.
Loss
Mechanism Test
The
Loss Mechanism Test diagnoses causes of
performance losses to direct engineers about how
to provide optimum maintenance of compressors.
Machine performances are improved to deliver
higher production flows.
Health
Index
A
Health Index for compressors has been invented.
This index is based on impeller geometry,
compressor speed and other selection criteria.
It is then compared with the process demands for which
it has been derived. The Health Index
identifies by a numerical value the suitability
of the design for the duty. It enables technical comparison of
alternative proposals to be evaluated and compared
so that the optimum is selected.
WHAT
HAS BEEN ACHIEVED
MSE
has developed a large database of compressors and
drivers and performed a rigorous analysis of the
operating data.
The
results of our analysis were presented in three
separate workshops that were organised at the
culmination of each Phase of the JIP project.
These
workshops enabled full interaction with the
engineers of the participating companies and
results of their machines were discussed in full.
JIP
MILESTONES
MSE
met all the JIP milestones and delivered to the
participating companies:
-
British Gas (HRL)
-
BP
-
ConocoPhillips
-
BG Group
-
ENI Lasmo
All
the participating companies were satisfied with
the outcome of the JIP. All of them requested
that the data collection and analysis of the
machines be extended for a further 12-month
period to enable additional data to be collected
and analysed to provide further input to the
database.
The
more machines in the database, the greater is
its effectiveness in identifying best designs
and rooting out operational problems.
DOWNLOAD
BROCHURE
|
