1994 INTERNATIONAL SYSTEM DYNAMICS CONFERENCE
Systems Thinking and its Influence on Operational Culture
JB Bentham AG de Visscher
Shell Nederland Raffinaderij B.V. Business Unit Offshore
Vondelingenweg 601 NAM
3196 KK Rotterdam/Pernis Schepersmaat 2
The Netherlands 9405 TA Assen
Fax: 31 10 431 3410 The Netherlands
Abstract
This paper reviews a recent period of change in the ‘ing envi of a large
oil refinery, and the role of systems thinking techniques in this development process. The focus is on
how the gradual adoption of these methods as part of normal working practices has contributed to
changes in attitudes, iour and per A ion is directed to the way the methods are
themselves proving to be a vehicle for cultural development, as well as being part of the toolkit used
to analyse specific issues.
The paper describes the way several techniques have been introduced, applied and combined. There
isa y on how these approaches have helped to induce a shift from a rather intimidating
fire-fighting lity to a wid d embrace of the quest for structural improvements. This shift
and the adoption of the techniques are, of course, mutually reii ing. What is of i
significance is the way the virtuous circle was initiated. It is concluded that the approach was
successful largely because it was consistent with the broader i of change
and compatible with fundamental aspects of the predominant organisational culture.
While this is an ongoing process of transformation, the paper introduces a framework to
conceptualise the insights which have emerged so far. It is intended that this will facilitate their
further application and hence support the broader learning process.
Organisational Learning, page 1
1994 INTERNATIONAL SYSTEM DYNAMICS CONFERENCE
Systems Thinking and Its Influence on Operational Culture
INTRODUCTION
As has been widely reported, a variety of systems thinking concepts, methods and processes have
taken root within the Royal Dutch/Shell Group of C ies in the past two decades. Indeed,
Shell managers and planners have at times been at the forefront of developments. The Group has
been particularly associated with Scenario Planning’ ”, Mental Modelling’ and, to a lesser extent,
Soft Systems Methodology’. Prompted and encouraged by formal corporate planning processes,
training and their use to facilitate dialogue on complex issues, these approaches are very familiar to
at least senior the ip They form a deeply influential aspect of
strategic learning within the Group, supporting, for example, decisions on major investments in
long-lived assets under ditions of often i
In the highly capital-intensive industries in which Shell operates, it is a simple truism that our
potential for success is founded on the quality of these decisions. Putting the assets in place,
however, is only the beginning of the story. They also have to be operated effectively, and linked
flexibly into their overall business environment. Without this, the benefits accrued through strategic
learning will be squandered. There is always going to be some difficulty, however, in establishing
effective learning processes in this operational area because of the large number of decisions,
people and activities it encompasses.
As line managers, we have recently been involved in a variety of initiatives to enhance
performance at the major Royal Dutch/Shell refinery at Pernis near Rotterdam. A cluster of
systems thinking techniques have been employed. As will be described, these methods have not
only brought the benefits expected from improved analysis, but also appear to have been associated
with a general shift in attitudes, perspectives and behaviour in the part of the organisation into
which they were introduced. The desired learning culture seems to be becoming established.
Looking back, there are a number of valuable lessons we feel can be drawn from this process, and
from comparing it with previous experiences at this site and elsewhere in the Manufacturing
Function of the company. We believe these insights will be appli to other ch
activities.
In this report we adopt a very broad view of what can fall under the general heading of “systems
thinking". For reasons which will become apparent, we include all techniques that encourage or
demand practiti to take an i d view of events. In other words, to balance a focus on
the individual events themselves with an appreciation of the relevant pattern of relationships
between them.
RECENT EXPERIENCE WITH INTRODUCING SYSTEMS THINKING TECHNIQUES IN
AN OPERATIONAL ENVIRONMENT
In the past few years, a number of different steps have been taken at the Pernis site to institute
performance enh: These have included, for example, the ISO9000 certification
of various parts of the business. It became clear, however, that one of the more formidable barriers
to improvement came from the unavoidable fact that a large number of different functional or
operational processes are involved in each of our business processes. With the aim of clarifying
and ly i ing such of “activities, the concepts and methods of Business
Process Analysis have been p d. ingly, the most i 1 of the various
pilot projects which were initiated was actually one of the more complex as it crossed several
departmental interfaces in very different parts of the business. It involved the commercial and
customer transactions of the Trading department, the operation of logistics facilities at the refinery
Organisational Learning, page 2
1994 INTERNATIONAL SYSTEM DYNAMICS CONFERENCE
by the Oil M dep and the dination and optimisation of these activities by the
‘ee ’and A
The team involved with this project began by developing a high-level overview of the logically
necessary actions required to accomplish the chosen business goal. There were similarities,
therefore, between this overview and the sort of conceptual models which result from applying Soft
Systems Methodology* (SSM), the main differences being that no explicit worldview or root
definition for the model were identified. Obviously, therefore, the potential value of investigating
the implications of alternative Advi or root itions was not explored. The more general
SSM process was introduced to the project group, but the relevance was not fully appreciated, and
the focus remained on the overview or "Functional Activity Model" as it came to be known.
Each of the logical activities of the main overview was broken down into sub-activities and sub-
models. It soon became clear, however, that an ion of ities in ever
greater levels of detail was Bone to emerge if this process was continued. To return focus to the
most relevant issues, ly prepared ive list of points of concern was
"projected" on to the activity model to highlight the areas where attention was most appropriate.
Eventually three linked activities at the second level of model detail were selected for deeper
analysis. Three teams performed detailed work-flow analysis in these areas, and then re-engineered,
d and impl h Appropriate attention was paid to ensuring that
quantitative feedback on the performance of the new processes was available and, to the
satisfaction of all involved, improvements were quite evident.
In addition, i i was d for this process of combining functional
activity models, points of concern and finally workflow re-engineering as a means of developing
insights into problem areas and then translating these into concrete actions at the operational level.
This approach has subsequently been applied in other projects.
At the same time, other hodologies were being introduced into the ‘ics and
department. From the world of Quality Management’, statistical process control (SPC) runcharting
was introduced to monitor deviations of actual plant, logistic and market operations from assump-
tions and instructions related to various time horizons. This was the first time such techniques had
been used at the site to focus on coordination and planning rather than individual well-defined
ducti ical and p ilistic methods were also adopted to analyse data and
relationships, with results used in simulations to support risk/reward analysis and decision-making.
A functional activity model for the was and ped into a Handb
detailing activities, responsibilities, information flows, business controls, document controls and
interfaces with partners in other areas of the business. This specific functional activity model was
then supplemented with more general conceptual models based on service provision and individual
fulfillment perspectives. In other words, a more complete SSM approach was adopted. These
conceptual models had a rich influence on self-perception within the department.
Causal-loop analysis and system dynamics concepts were also introduced. Initially application was
confined to exploring physical but soon of the d were includi
the impact of such variables as “anxiety” and “credibility” in their analyses, and taking these into
account in their decisions and activities.
One of the most exciting facets of these developments was the way in which an initial lack of
interest, scepticism, and even resistance were gradually replaced by enthusiasm, imaginative
application and curiosity about other “exotic” techniques. This was also accompanied by a shift in
the climate of attitudes and behaviour. A calmer and more systematic approach began to be
adopted in routine activities, and there was a widespread embrace of the quest for structural
enhancem nt. There was also an imp level of de di of the mutual roles,
ies and ii i of staff and ional staff, an area previously
characterised by information blocks and excessive intervention. As an example of the progress, the
Organisational Learning, page 3
1994 INTERNATIONAL SYSTEM DYNAMICS CONFERENCE
following is an extract from a ion on the SPC phi prepared by bers of the
department and given to their in other
“Management understanding" means “Better use of Management"
Employees: Duty to deliver information in understandable form.
Management: Duty to keep “hands off" if process understood and
demonstrably under control.
Besides anecdotal evidence related to specific examples as above, the change in climate was also
apparent from the results of an ry ire. The average resp to the i
most related to the subject of this paper are shown below. There is a clear appreciation that
something has changed, and hopefully for
the better!
Comparison with
year
It is almost self-evident that there should worte bate
be a connection between adopting
systems thinking techniques and a shift The ollmate tn the
: depariment is positive
towards a more systematic and structured
view of the environment. It is not
My work is interesting
surprising that the change in behaviour and challenging
goes hand in hand with the change in
perspective, and that they are mutually \1am encouraged to make
reinforcing. A more complex issue to structural Improvements in
4 ae rk
understand, however, is how this virtuous are
circle was able to be initiated in the first i 7 i
shes ws have confidence in the
place. In a very similar department with way the department is
a similar profile of employees at another seveloping
Shell’ refinery, ‘the, ‘very same, line The department contributes to
manager had attempted to initiate the structural enhancement of
same kind of change, but with relatively sp Duslnwes eattorranes
little impact. To really get to grips with
our experience, therefore, at least the
ig i need to be addi
1. Why has the ics and Scheduling d proved to be parti ly receptive to these
changes?
2. Why were the changes embraced at this site and not in a very similar situation elsewhere?
3. Which systems approaches were most enth ically adopted and why?
4, How did the factors build on each other, providing a path to ing and
progressively more abstract and powerful perspectives?
In any kind of complex situation, a wide variety of alternative explanations is. possible.
Nevertheless, in what follows we have tried to encapsulate the quality of our own specific
experience in a consistent way which draws heavily on frameworks derived from the literature.
Organisational Learning, page 4
1994 INTERNATIONAL SYSTEM DYNAMICS CONFERENCE
THE DYNAMICS OF CULTURAL CHANGE
On the basis of his extensive work on the culture of isati Schein? ibes the
of change in terms of processes of "Unfreezing", "Cognitive Restructuring" and "Refreezing". To
unfreeze the existing culture there must be sufficient gap between performance and expectations to
produce anxiety or guilt. This disconfirming data establishes the need for new learning and a re-
definition of core assumptions. If and when such changes lead to performance enhancement, their
validity becomes accepted and they will gradually become elements in a new cultural pattern.
As Schein and others" ™ have pointed out, however, it is quite possible for members of an
organisation to deny the relevance or validity of di ‘ing data for very i le periods,
for example by a process of rationalisation which jally reduces i In addition, the
anxiety by the perf gap can letely arrest the required learning process
because people become too insecure to conduct the experiments needed to generate fresh insights.
The treadmill of "doing more of the same" while rationalising unsatisfactory performance becomes,
therefore, a substitute for learning and adaptation. Making use of the "Eroding Goals" system
archetype described by Senge", the dynamics of
this process are illustrated below.
For change to take place successfully, therefore, “Eroding Goals” structure in Change Processes
Schein’ points out that there must be a climate of
psychological security in which change can be
“vette Fra inant
contemplated without fear, as well as the ko *, f* *y
of ning data to
elon
important fixed goals and ideals. Finally, there cis satin Seetnonere
must be concrete experimentation based on trial ( = a
and error or the adoption of new role models. Gap te Coane oh Annieny
E+ ciscanance ——t ane ‘inseesty
To effectively support or promote change, *
ion should be — roves
“Secintion avonaiae gee
consciously directed into these three areas, as
shown below. Ne
en
‘ ( Frash wnaighla \
vaa
Jong-teem Cognitive
pertormance gusly Restructuring Experimentation
Me tn
i
Gap ——s pistonance fang insecunty
[- \.
WY
Organisational Learning, page 5
1994 INTERNATIONAL SYSTEM DYNAMICS CONFERENCE
One of the reasons for the relatively disappointing response to attempts to initiate similar changes
at another site can be identified from this structure. A long period of business difficulties and
general recession had resulted in a degree of anxiety and suspicion which undermined
Il security and, fore, the forces for change. In contrast, the situation at Pernis was
less extreme and there was considerable management action to promote a climate of trust.
TOWARDS THE LEARNING CULTURE
To get to grips ‘with the changes we believe are taking place, we first need to Gevelop a description
of the original situation. In this, we again follow the lead of Schein’ in characterising
organisational culture according to a number of key dimensions, summarised in a self-consistent
and reinforcing pattern of underlying assumptions. Bentham? has given a detailed account of
applying this type of analysis to describe the assumptions prevalent in the Central Office of the
Manufacturing Function in the 1980's. Not surprisingly given the degree of personnel exchange, the
pattern which emerges for Pernis shows many similarities with that developed for the central
manufacturing function. The most relevant di: are the hi is on concrete
experience as the fundamental source of individual credibility, and the focus of attention on
immediately current events.
These should not be surprising ch istics for an "O i culture!
The pattern of cultural ions in the Pernis O: i E
Wa ure a major production
facility in @ tough
auvironment, Our
tnigsion Ip to wxplolt our
resources to fight the
ongoing battle againat
the forces of chaos on all
tronte.
Individuals have different
innate potential tor dealing
with complex technichal.
operational and
We are a traditiunal “Inwih and Wiedom* reside
ilerarchy. and 1 iu with those who have the
most exprriencersnniority,
+——+ _ and who are capable of | ¢___y inc umanaperent
‘Challenge is. ReceDy ie. bur asnertively expressing and issues. This can be ideniilied
nay fear tam wile. ba acting upon their private ye es
stabs hh resihiy kauwiedge. slog whatever sary ane besouht os
supparting evidence ia hough Oa ss
invallabia
Rol it delineated External retations and those
er ars well ceuted ted The strength of the organisation isaleauncdagnuncnts ce’ very:
and normally considered ¢@————» : > bet inaun departments are von
ae hase potential tor loss of “face”
Se uate Hard-won credibility should
however. te coutinitally eon cecaly Se
challenge thoxe who have
vigorous response to Inreate
or mutually understood
non-conteontation.
not (yo!) ostabilisned
Organisational Learning, page 6
1994 INTERNATIONAL SYSTEM DYNAMICS CONFERENCE
It is valuable to compare these characteristics with those of a "Learning Culture" as described by
Schein’. The dimensions where there appear to be significant differences are summarised in the
table below.
DEVIATION BETWEEN CHARACTERISTICS OF "LEARNING CULTURE" (9%),
AND “REFINERY OPERATIONAL CULTURE" (X)
Nature of Human Nature
Humans basically evil Humans basically good
.
Human nature fixed Human nature mutable
.
Nature of Human Relationshi
Groupism Individualism
°
Authoritative‘qatemalistic. RaMeegedirpeticcaave
x .
Nature of Time
Past onvented Present oriented Near-tuture oriented
x
Short time units Medium time units Long time units
x °
Information and Communication
Low level of connectivity Fully connected
x °
Subculturat Uniformity Versus Diversity
High unitormity High diversity
x e
Task Versus Relationship Orientation
Primarily task oriented Task and relationship Primarily relationship
oriented oriented
x e
Linear Versus Systemic Field Logic
Linear thinking
To simplify matters, we distil these relevant characteristics into two summary dimensions.
The first of these disti from ii ‘ion, and thus echoes something of the
flavour of the work of Bohm’?. The second makes a distinction between a natural tendency to
interpret events from a mechanistic or objective perspective as opposed to a humanistic or
subjective worldview. Based on the i ised above, and p given the
monochronic view on the nature of time and the emphasis on the present moment and short time-
scales, we would describe the lying refinery p p as d". There is also a clear
y towards the “mechanistic” ori
Organisational Learning, page 7
1994 INTERNATIONAL SYSTEM DYNAMICS CONFERENCE
In contrast, the learning culture is relatively integrated on one di ion and more
between the poles on the other. The desired and, to some extent, experienced shift in cultural
orientation can, therefore, be summarised via the illustration below.
Desired shift in Underlying Orientation
Frag Fh enemy —| integrated
Perepestive Perspective
Mechanistic Humanistic
Objective —o——> —. — Subjective
Orestation F Orientation
These dimensions are intended to summarise the general culture of a group as it is related to
organisational learning. It is also of interest to note, however, a degree of consistency with
descriptive ft ‘ks applied to ch ise individual and learning style.
Lewis and Margerison* have analysed the strong correlation between the factors used to describe
Jungian archetypes and the Kolb/Plovnick learning style model. We hypothesise that our
fragmented/integrated dimension is related to the and axes of
these frameworks, and that our mechanistic/humanistic axis is related to the thinking/feeling and
active/reflective di i Given a ip between cultural assumptions and individual
style, it would seem likely therefore that a istic culture would tend to be
associated with individuals having ing/ and acti inking ori i Using standard
questionnaires to assess individual style and preferences, we have indeed observed this tendency in
our admittedly limited experience. In addition, these inventories give an indication of diversity and
balance within a group, which are also considered to be important factors in organisational
learning. We will certainly consider using this approach in future to assist in cultural analyses.
Using this framework, it is now possible to offer an ion for the parti receptiveness of
the Economics and Scheduling department to change. The cultural pattern illustrated previously
describes the general orientation of the refinery, but within this there are a variety of nuances
ishing different sub-cul . The nature of coordination and planning activities, for
example, p a more polychronic app ion of the flow of time, a greater focus on the near
future and a greater emphasis on relationships as well as tasks. Being closely involved with the
very different worlds of refinery operations and trading, the members of the department also have a
deep awareness of the very different perspectives which can be brought to bear on the same events
and activities. In other words, the sub-culture found in this type of department is already one small
step closer to the learning orientation than the bulk of the organisation and is therefore more likely
to be responsive to the changes introduced.
THE CONTRIBUTION OF SYSTEMS THINKING TECHNIQUES
As indicated previously, the re-structuring of assumptions which forms part of the change process
is driven by fresh insights. Systems methods can eoyiensly play a role in this. In circumstances
where the underlying perspective is pred ly d, experimenting with these methods
can lead to a new appreciation of relationships between events previously treated in isolation,
placing them in a context of causal loops and meaningful patterns developing over time. In terms
of the framework which has been developed, this will tend to promote a shift of underlying
assumptions in the direction of a more integrated perspective. Mental mapping and soft systems
Organisational Learning, page 8
1994 INTERNATIONAL SYSTEM DYNAMICS CONFERENCE
dology also focus ion on the i of p ive on interpretation and action, and
can encourage a shift towards a more balanced position in an isation with a
mechanistic/objective orientation.
To promote these changes, however, requires that the approaches are applied widely and regularly
over a reasonable period. They need, therefore, to be accepted by the majority of individuals as
relevant working techniques which justify the time and effort involved in learning how to use
them. Half-hearted application is unlikely to be successful. The reaction of an organisation to
unfamiliar methods and ideas is often, however, analagous to the reaction of the immune system to
unfamiliar bacteria. Our experience has taught us that preventing premature rejection requires
careful matching of the initial approach with the existing caltore: In an operational environment
with a strong ori ion towards istic and ives, for example, the value
of mental mapping and soft system methodologies is not likely to be immediately recognised. This
was certainly our experience in trying to introduce systems thinking via such methods at both sites.
In contrast, however, if people perceive the relevance of an approach, they are more likely to apply
it willingly and it is more likely to have a positive impact. In turn, this builds confidence and
unfreezes perspectives, enabling further approaches to be introduced, and so on. This dynamic is,
therefore, an engine of change contributing to the generation of fresh insights and the reduction of
anxiety levels, as illustrated below.
Fresh
insights
Kick-Starting the Engine of Change Z
legree of
experien tation
“po
Willingness to
try out new approaches ji Success of
ud + oP Engine application
- of of new methods
change
Recognition of
approach relevance
+ =
Anxiety and
insecurity
Compatability
of new approach
with existing
perspective yCa00 H
Organisational Learning, page 9
1994 INTERNATIONAL SYSTEM DYNAMICS CONFERENCE
Although not normally considered a systems thinking the i ion of
process control methods provided the required "kick-start" for the Economics and Scheduling
department at Pemis. The accent on measured, concrete data was compatible with the
isti culture, while it encouraged the first steps towards an
by highlighti trends, patterns and other systems properties. There was the
further advantage that the style of data presentation was familiar to those working in a process
industry. In addition, SPC made disconfirming data on performance highly visible in a non-
ing manner, ising the per gap without undermining psychological security.
Similarly, the construction of a functional activity model was directly appealing within this
environment.
These first steps paved the way towards experi ion with a of ii
conceptually complex and abstract methods which would have been considered alien and irrelevant
in the original climate.
Conceptual Steps in Responding to Complex Situations
Appreciation of
Generic Structures
ippreciation
ternative
cognitwe Systems
Appreciation
of Functional
Systems
Trend
Perspective
Event
Perspective
A similar progression from integrative simplicity to integrative complexity is described in theories
of cognitive style related to the development of individual personality’. In this case we have
observed development from a description of the world in terms of individual events through to an
iation of system arch 1° and generic
CONCLUSIONS
We summarise the lessons from our experiences by means of the following recommended process
for introducing systems thinking techniques as catalysts for behavioural change in an organisation.
This emphasises the broader context of change-management previously described and the need to
match the initial approach with existing characteristics and perspectives. We have found that the
framework of cultural analysis briefly outlined in the text can be usefully applied to achieve this.
When introduced appropriately in this way,
Organisational Learning, page 10
1994 INTERNATIONAL SYSTEM DYNAMICS CONFERENCE
we believe a wide spectrum of systems thinking methodologies can contribute to the desired
process of organisational development.
Appreciate how [——* Monitor adoption
Ararat syne tig aimee
and identity desired =a contribute to ee
charigie desired changes development
\ NX Take control
Identity action
methods which
Identify will be rapidly |
(sub)groups ——~—————> _ valued, and
inclined'susceptible ‘suitable initial
to change applications
introduce and
apply Systems I~
Thinking methods
ieentty piers
relovant discontirming Create a Ensure
data, ond ensure it climate ot ~~ sutticient time oe
is approciated psychological and “space" allocated feedback ‘a
security to application their value
\ Identity
‘opportunities to
extend appreciation
and application
REFERENCES
1. Aronson, Elliot, The Social Animal (5th edition). New York: W.H. Freeman & Co., 1988.
2. Bentham, Jeremy B., METAPHORMOSIS: Transforming Structured Mi h into Practical
Management Tools, Master's Thesis, Sloan School of Management, M.I.T., June 1991.
3. Bohm, David, Wholeness and the Implicate Order, New York, Routledge, Chapman and Hall,
1980.
4. Checkland, Peter & Scholes, James, Soft Systems Methodology in Action. Chichester: John
Wiley & Sons, Ltd., 1990.
5. de Geus, Arie P., "Planning as Learning", Harvard Business Review, Vol. 66, No. 2,
March/April 1998, pp. 70-74.
6. Goldstein, Kenneth M. and Black Sheldon, "Th ical to itive style",
P. lity, Theory M and R h (ed) Fay Fransella, London, Methuen and Co.,
1981.
7. Grant, E.L., and Heavenworth R.S., Statistical Quality Control, 6th ed., New York: McGraw-
Hill 1988.
8. Lewis, Ralph & Margerison, Charles, "Mapping M ial Styles", Ir ional Journal of
Manpower, Vol. 2, No. 1, pp. 1-24.
9. Schein, Edgar H., Or izational Culture and Leadership, 2nd ed., San Fransisco: Jossey-
Bass, Inc., 1992.
10. Senge, Peter, The Fifth Discipline, New York: Doubleday, 1990.
Organisational Learning, page 11
1994 INTERNATIONAL SYSTEM DYNAMICS CONFERENCE
11. Wack, Pierre, "Scenarios: Shooting the Rapids", Harvard Business Review,
November/December 1985
12. Wack, Pierre, "Scenarios: Uncharted Waters Ahead", Harvard Business Review,
September/October, 1985
13. Weick, Karl E., The Social Psychology of Organizing, 2nd ed., New York: Random House,
1979.
Organisational Learning, page 12
