REDUCE GRAPHICS TO IMPROVE OPERATORS’ PROCESS

OVERVIEW

Ruurd N. Pikaar Eur.Erg.

ErgoS Engineering & Ergonomics

The Netherlands

Process operators control and supervise complex processes using several

screens with process information. To enable the operator to do an adequate job,
instrumentation engineers and process control engineers need to address several
interrelated topics such as: 1) console and instrumentation design, 2) process
graphics, 3) human-computer interaction, and 4) alarm management. The
primary focus of this paper is on the design of process graphics, which cannot
be done without addressing the other topics. However, in process control
upgrade projects, one usually aims for a 1:1 conversion of the existing graphics.
On user request frequently larger and more displays are introduced at the
workplace. It is shown that a 1:1 conversion is not a good approach. It would be
better to analyse the old graphics for ease of use and task performance first.
Applying ergonomic guidelines for information display, simplifies existing
graphics considerably. In turn, this reduces the number of process graphics
needed, meanwhile improving process overview, i.e. simplifying the operator
process supervision task.

Ergonomics and Engineering

This paper concerns control system upgrade projects and related graphics (re)design. Based
on ergonomic guidelines for information display, designing graphics for specific types of
processes and control systems is discussed. First a background on the role of the ergonomist
or human factors engineer (both terms will be used as synonyms) will be given. The contri-
bution of ergonomics in engineering projects is limited compared to technical engineering.
Projects run through several phases, often starting with a feasibility study, via analysis and
design phases, to detailed engineering and implementation. Usually, each phase ends (or
starts) with a go-no go decision based on formal documents. The ergonomist is never the
only actor in a project and may not be involved in all project phases. The aim ergonomics is
to help optimizing the work system, by applying a user centred approach. This requires a
joint design of both social and technical aspects of the system. Job design, operator
workload, control centre and workplace layout, instrumentation, information display,
environment and many more topics have to be addressed. The human factors professional
may not have much background in process control or other engineering sciences. Therefore,
he/she first of all relies on methodology, i.e. a systematic design approach including a task-
and situation analysis (Pikaar, 2007 & 2008). He/she tries to get insight in the relationships
between relevant human factors, thus touching on work of other disciplines within a project.
For example, the process control engineer may be responsible for the contract with the
instrumentation vendor, deciding on workplace equipment (“we need modern wide
screens..”), and graphics (“you may use the process flow diagrams as a starting point..”).
Operations may think otherwise. Civil engineering dictate workplace location because of
cabling requirements. In some cases we noticed that nobody took responsibility for the
design of operator workplaces. The ergonomist may fill in the gaps between technical
engineering disciplines. In addition, standards may be helpful to convince a project team to
include a human factors approach in the upgrade project. Industry in general is sensitive to
standardization. Standards as such the series ISO11064 Ergonomic design of control centers,
provide useful design requirements, include mandatory requirements on the design process.
Parallel to the development of ISO11064 series, the International Instruments Users’
Association WIB (the Hague) published practice based ergonomic guidelines for process
control room design (see Pikaar et.al., 1998a, 1998b).

Methodology

This paper is based on several case studies, two of which are presented in some detail in the
section Process graphics upgrade projects. Case studies have a methodological problem.
Statistically valid comparisons of several projects cannot be made. A project is not carried
out twice, i.e. with or without ergonomics. Also, it is rarely possible to redo a detailed task-
and situation analysis after a project is completed (in practice the project will not supply
funding for it). The author strongly believes that experiences with human factors engineering
in industrial settings with many practical constraints, should be reported in ergonomics
literature notwithstanding methodological problems. Publication is needed to bridge the gap
between scientific research and practice (Pikaar, 2008). In the case of graphics design, new
ideas may be developed and tested against old presentation formats in a laboratory setting,
with student volunteers instead of process technicians and life processes (measuring response
times, etc.). The author did so during his master study, many years ago and so may have
done other researchers. Still, no scientific data can be found on what is to be considered too
much, too little, or just the right amount of content (i.e. the number of controlled variables)
on a graphic. And this is exactly the first questions of a process control engineer. Feedback
from actual projects should be welcomed. The systems design approach of all our cases has
been in line with the process requirements of ISO11064-Part 1. The systems design
approach consisted of 1.) task- and situation analysis, 2.) development of task related design
requirements, 3.) implementation, and included 4.) user participation. During a situation
analysis an inventory is made on the process characteristics, which is related to type of
process industry (for example distribution systems or oil movement require other types of
process control then refinery processes). Also an inventory of capabilities and limitations of
the instrumentation system is needed. Next an ergonomics style guide is developed for text,
characters, symbols and standard equipment (design requirement phase). During this phase,
background on guidelines for information display is presented to process operations
representatives. In the end they will be responsible for developing and approving graphics.
First new graphics may be developed by ergonomists and discussed with operations,
instrumentation engineers and vendor. Next, operators gradually take over the development,
assisted by the ergonomist with comments, suggestions, check on correct use of the style
guide and so on.
Graphics and guidelines

To enable process operators to do an adequate job, control room designers need to address at
least four interrelated topics: 1.) console or workplace 2.) process graphics, 3.) human-
computer interaction, and 4.) alarm management. The focus of this paper is on process
graphics, however the other topics are needed as a design context or operation philosophy.

Console design
Workplace layout includes the instrumentation and equipment needed and how it is
organized. The number of screens on the desk and elsewhere (such as large displays for
overview or cctv) determines workplace sizes, possibly overview of information sources, and
viewing distances. To design graphics, we need to know about viewing distances. Legibility
of important text is guaranteed at a maximum of 200 x character height (capital). For
secondary text this may be 250 x character height. At a viewing distance of 1 m, this results
into 5 mm, respectively 4 mm character height (ISO 11064). Legibility guidelines are based
on the critical detail an average person (visual acuity: 1) is able to distinguish, i.e. one minute
of arc or approximately 1 mm at 3 m distance. For engineers, this guideline is straight
forward. Confusing will be sources that indicate different approaches with other outcomes,
amongst others based on the diagonal size or width of screens (Hénique et.al, 2007). The
simple question of viewing distances also leads to different points of view. Mainly in
Europe, you will find workplace designs based on the operator to be able to read information
from up to 4 displays in a (curved) row. Typical viewing distances will be 850 – 1000 mm.
USA based guidelines (for example ASM, 2008) assume the operator to be working with one
screen out of many at a time, at (very) short viewing distances of 600 mm. Anyhow, console
design influences character sizes, screen sizes and graphics layout.

Human computer interaction

Graphics are part of the operator interface of the process control system. The selection of
graphics, detailed process data (controllers), alarm overview and other pictures is governed
by interaction software. Depending on how it is organized, interaction requires space on the
screen, and it is usually repeated on each screen. For example, if you want to have a one
touch access to each graphic, a button is needed for each graphic. Other examples of data
repetition are time and date, or the last three alarm lines, being shown on every display.
Interaction design is vendor dependent. The number of graphics has an impact on interaction
needed. Of course this also relates to the number of screens on the console, ease of selecting
and finding data, ease of switching between graphics and so on. Ergonomists have the tools
to review and optimize software (Mulder, 2006).

Alarm management and Process Overview

Complex processes require tools to present an overall view of the processes to be supervised
and the status of off-normal situations (alarms). Nowadays, the use of CCTV systems
(camera observation of critical process areas) is also introduced. Many seem to favour the
use of large videowalls or display panels in the control room. Two rather different points of
view can be found in Hénique et.al. (2008) and De Groot (2006). Hénique narrates about
available technologies, without questioning the usefulness in the control room environment.
De Groot takes a task oriented approach and developed powerful process overview displays
for several applications. For multi operator situations, such displays are permanently
displayed on large, wide screen, TFT-screens. As a consequence, process overview data will
not be needed on the console based graphics. Off-normal process information (commonly
indicated as alarms) adds to the overview, i.e. improving operator awareness of the process
conditions. In many cases an off-normal overview is presented on a separate screen. Recent
developments can be found in the ASM Guidelines for effective display design (Bullemer
et.al., 2008). Although technology pushes cctv-pictures on large screens into the control
room, there is a lack of knowledge on how best to display cctv-pictures, as Wood (2007)
shows in an overview article.

Graphics design
As indicated above, graphics design requires insight in workplace characteristics and process
operations philosophy. Guidelines for process graphics design can be found in Bullemer
et.al. (2008) and Pikaar et.al. (1998b). They can be summarized by one design rule “keep it
simple” or, in case of existing graphics, “simplify as much as possible”, and one content rule
“present only task related information”. Both rules can be detailed further and differ per
industry (on the task level) and per instrumentation vendor (on a style guide level).

Figure 1. Original graphic (one full screen).

Figure 2. Simplified tankage graphic (half screen graphic).

The example illustrates the general idea. Figure 1 shows the original graphic. Figure 2 shows
the simplified one. Changes include filled in shapes of tanks (no border lines) and reducing
the process line-up to its actual functionality (instead of the layout in the field).

Process Graphics Upgrade Projects

Case study 1 – Food industry

Situation: three local control rooms have been combined into a central control room, mean-
while upgrading the old instrumentation systems and adding a new process line. The proces-
ses are mainly batch oriented. Operators are responsible for control room tasks as well as
field tasks. They walk in and out the control room. Once at the console, they wish to have a
quick overview of the process status (because they have been in the field). Therefore, each
workplace had up to 6 DCS screens and several dedicated screens.

Design approach: after a situation analysis in each control room, a working group of operator
representatives was formed. The members got an introduction in graphics design. On the
basis of several examples for each plant (local control room), the idea to introduce new
graphics was accepted. One senior operator and the ergonomist redesigned the 121 graphics.
This resulted into a total of 40 new graphics plus one overview graphic per plant.

Result: because of the work organizational impact, i.e. the change over to centralized control,
a strategy for change was needed. In this case all old graphics were converted 1:1 for the new
I/A System. In addition the 40 new graphics were introduced as an intermediate layer of
pictures between the permanent process overview and the converted old graphics. All data of
the existing graphics could be found on this layer. Figure 3 and 4 give an example of a
redesigned (originally full screen) graphic.

Figure 3. One of the original graphics.

Figure 4. Redesigned graphic.

After implementation and a period of time during which the operators could get used to the
new displays, the old graphics now seem to have been abandoned by almost all users.

Case study 2 – Chemical plant

Situation: at this plant the three main processing lines are continuous, except for one batch
processing step half way. An important process control feature is to plan carefully the multi
batch processing. There is one control room operator and there are four field operators. The
plant is to be extended by a 4th line, including an estimated 55 new graphics.

Design approach: the situation analysis in the existing control room showed that the operator
has a high workload. The job can be done, however it was questionable whether the operator
would also be able to handle an additional 4th line. Finding process control data was
elaborate, in particular because each of four batch reactors per line was presented on a
separate graphic. It was decided to redesign existing graphics for implementation in the new
production line. Next management would decide on further steps regarding the existing lines.
One operator per operator team participated in a working group. A workshop on ergonomic
design guidelines was organized. Within a few weeks all basics were thoroughly understood
and the working group took over redesign work almost completely and successfully from the
ergonomist. A DCS vendor engineer also participated and helped developing the style guide.

Results: instead of the estimated 55 graphics at a cost of € 55.000, 20 graphics were

developed including a graphic containing all 4 batch reactors. Static graphics design was
done by the working group members during quiet operation hours (evening/night). Next the
DCS vendor implemented the graphics, and the operators work group did factory acceptance
tests. Based on enthusiasm of the operators and the relatively low costs, management decided
to have all graphics redesigned, while upgrading the old system. The existing 3 production
lines including utilities showed a reduction from 200 to 70 graphics. One additional
argument played a major role. Because of the extension to 4 production lines, the control
room was planned to expand from 1 to 2 operators. However, it was also estimated that a
reduction of operator workload could be achieved by improving information display. As a
result one operator now handles all processes during normal process conditions. Note: figure
1 and 2 have been taken from this project.

Other Cases
In both previous cases a full upgrade of the process graphics initially was not intended, but it
came out as a result of a convincing partial conversion. The authors company participated in
other conversion projects in power generation, chemical industries, and oil (exploration)
industries. The projects included major instrumentation systems (Yokogawa, Invensys,
Honeywell, Siemens, Arreva, ABB). In all cases, guidelines or style guides have been
developed as well as graphic redesign examples. Usually the project owner then took over
further development and implementation. Detailed outcomes are not always fully known to
the author, or may not be published in detail. However based on these experiences there is a
high probability that a human factors knowledge driven upgrade of graphics reduces the
number of process graphic for almost any type of process industry by at least 50 %.

Discussion

Every 10 to 15 years process control systems are upgraded. Based on assumed relatively low
costs, engineering or project management aims for a 1:1 conversion of the existing graphics.
The instrumentation vendors developed powerful software tools to enable this. Based on
several recent cases, a 1:1 conversion is not considered to be a good approach. There are
many reasons for this statement, amongst others:

• Improved display technology, in particular higher resolutions and wide screens, enable
better representations of symbols, process equipment, indicators, etc. A 1:1 conversion
doesn't use these advantages. In upgrade projects, the old graphics have been developed
15 to 20 years ago. An ergonomic review of graphics usually reveals cluttered and ill
structured data. Ill structured data doesn’t assist the operator in acquiring a good process
overview, which is the main task of an operator. Unit shut down or production losses
have been reported during task analyses, because operators missed vital process
information on disturbed graphics (this is usually not reported or it is not allowed to
publish about it).

• Old technology graphics can be considerably simplified, applying a limited number of

straight forward ergonomic guidelines on information display. Both cases have shown up
to 67% reduction of the number of process graphics; other cases show an average of 50%
reduction. Except the obvious improvement of the operator task, this also represents a
significant cost reduction (less graphics, less development costs, less maintenance).

• In addition a permanent process overview display may be a powerful tool. An overview

typically consist of key performance indicators and most important process variables. A
detailed task analyis will reveal its possible content. For example, a waste incinerator
including flue gas treatment can be summarized in 20 – 25 process variables, according
to experiences process operators. Operators know a permanent visible display by heart;
therefore no, or very limited, static process details are needed on such a display.

• Usually, content (processes and process equipment) changes over the years, needing an
adaptation of existing graphics during or after conversion. Practical experiences of 1:1
conversion projects indicate that it is rather seldom that the conversion succeeds without
unexpected problems due to process changes that may not have document in full.

• Graphics and interaction design have an impact on console design and vice versa. There
is a major impact on operator workload. If it is difficult for one operator to get a good
overview of the process status, the need may be felt to have two operator for the same
 process control job. If a task analysis shows that this is due to badly designed graphics,
more cost effective solutions can be implemented (refer to the second case study).

Conclusion

Instrument upgrade project teams should consider human factors inspired graphics redesign
instead of a 1:1 conversion. It is highly recommended to embark on the suggested approach
on the basis of an operator task analysis / situation analysis and a review of existing graphics.
Simplifying graphics or reducing graphics to their process functionality, reduces the number
of graphics considerably, i.e. by 50% or more. In turn this makes it easier for the operator to
get the process data he needs to do an adequate job. Thus, reducing graphics improves
operators’ process overview. Do not copy 15 year old graphics onto new technology.
However, you may consider to copy the 25 year old idea of permanent (conventional)
process and alarm status panels on modern technology displays. The human factors
profession has the tools and knowledge to investigate workload issues, workplace design,
information design, and knows how to organize user participation.

References

Bullemer, P. et.al. (2008), ASM Consortium Guidelines – Effective Operator Display Design
(Honeywell International Inc./ASM Consortium).
Groot, N. de, Pikaar, R.N. (2006), Videowall Information Design: useless and useful
applications. In: Meeting Diversity in Ergonomics, Proceedings of IEA2006 (Elsevier,
Amsterdam).
Hénique, E., S. Lindegaard, B. Hunt (2007), Design of Large and Complex Display Systems.
In: Ivergard, T. & B. Hunt, Handbook of control room design and ergonomics (CRC
Press, Taylor & Francis Group, London), 83-130.
ISO 11064 (1998-2007), Ergonomic Design of Control Centres – multi part standard.
International Organization for Standardization.
Mulder, E. (2007), An Integral Approach to make Software Work , In: R.N. Pikaar, A.E.P.
Koningsveld, & P.J.M. Settels (eds.), Meeting Diversity in Ergonomics, (Elsevier,
Amsterdam), 211-228.
Pikaar R.N. et.al. 1998a, Ergonomics in Process Control Rooms, Part 1: Engineering
Guideline. (WIB International Instrument Users' Association, The Hague).
Pikaar, R.N. et.al. 1998b, Ergonomics in Process Control Rooms, Part 2: Design Guideline,
(WIB International Instrument Users' Association, The Hague).
Pikaar R.N. (2007), New challenges: Ergonomics in Engineering Projects. In: R.N. Pikaar,
A.E.P. Koningsveld, & P.J.M. Settels (eds.), Meeting Diversity in Ergonomics,
(Elsevier, Amsterdam), 29-64.
Pikaar R.N. (2008), Ergonomics in Engineering Projects - How to achieve a booming
business. In: L. Sznelwar, F. Mascia and U. Montedo (Editors), Human Factors in
Organizational Design and Management – IX.
Wood, J. (2007), CCTV Ergonomics: Case Studies and Practical Guidance. In: R.N. Pikaar,
A.E.P. Koningsveld, & P.J.M. Settels (eds.), Meeting Diversity in Ergonomics,
(Elsevier, Amsterdam), 271-287.