Available in English, French and German

BEST
European Foundation
for the Improvement of
Living and Working Conditions

CONTINUOUS
SHIFT SYSTEMS

Bulletin of European Studies on Time

Number 11

Bulletin d’études européennes sur le temps

Bulletin für europäische Zeitstudien
ISSN 1017-4877
 European Foundation
for the Improvement of
Living and Working Conditions

CONTINUOUS SHIFT
SYSTEMS
edited by
Alexander Wedderburn

Bulletin of European Studies on Time

Bulletin d’études européennes sur le temps
Bulletin für europäische Zeitstudien
Original Language: English

Luxembourg: Office for Official Publications of

the European Communities, 1998

ISSN 1017-4877

Any items included in a BEST Bulletin may be reproduced without further

permission if the source is acknowledged. If you wish to receive either individual
contributions or more copies of BEST please write to the Foundation.

Printed in Ireland

EUROPEAN FOUNDATION
FOR THE IMPROVEMENT OF LIVING AND WORKING CONDITIONS
Wyattville Road, Loughlinstown, Co. Dublin, Ireland
Tel: (+353) 1 204 3100 Fax: (+353) 1 282 6456
Contents

CHAPTER 1: INTRODUCTION 6
The mathematics of shift systems 7
Guidelines for a good shift system 8

CHAPTER 2: PREVALENCE (FREQUENCY) OF CONTINUOUS

SYSTEMS 9
Italy 9
Belgium 9
Germany 11
Netherlands 11
France 12
United Kingdom 14

CHAPTER 3: SECTORS WITH CONTINUOUS SYSTEMS 17

Introduction 17
Changes and Trends 17
Textiles 17
Bridging shifts 19
Metallurgy 19
Chemical sector 20
Electricity-generating sector 22
Some general information about other work sectors 23
Transport 24

CHAPTER 4: WAYS OF MANNING CONTINUOUS SYSTEMS 25

Introduction 25
Four crews : rapid or slow rotating, or fixed 25
Five crews 27
Six crews 33
Twelve-hour shifts 35
Poor systems 37

CHAPTER 5: UNEVENLY MANNED CONTINUOUS SYSTEMS 40

Telephone call centres 40
Underlying principles 42

CHAPTER 6: CONCLUSIONS 43
References 44

3
 List of tables and figures

Table 2.1 : Population with a job: shiftwork, irregular hours, by economic sector:
Belgium (1994) 10
Table 2.2 : Frequency of shiftwork by size of enterprise (in %) 11
Table 2.3 : Distribution of shiftworkers according to shift systems in
France (1990) 12
Figure 2.1 : Continuous shiftwork in different sectors in France (1990) 13
Table 2.4 : Percentage of employees working shifts: UK 14
Table 2.5 : Percentages of men and women in different types of shiftwork :
UK, 1992 vs. -1994 14
Table 2.6: 3-shift workers in various industrial sectors (in thousands): UK
(1992 vs. 1994) 16
Table 2.7 : Continental shiftworkers in various industrial sectors (in thousands) :
UK (1992 vs. 1994) 16
Table 3.1 : Shifts in the textile industry in Belgium 18
Table 3.2 : Continuous shift system in a steel company in Northern Italy 20
Table 3.3: 12-hour 5-crew system, with 4 consecutive shifts:
UK chemical industry 21
Table 3.4 : Variety in shifts in the Belgian chemical industry 21
Table 3.5 : Six-crew shift rota in an electric plant in Italy 23
Table 3.6 : Former five-crew shift rota in an electric plant in Italy 23
Table 4.1 : Changes in shift patterns in a Belgian packaging company 25
Table 4.2 : Continuous 4-shift system in the German chemical industry 26
Table 4.3 : Continuous 4-shift system in a Swiss oil refinery 27
Table 4.4 : Change from 4- to 5-shift system in the paper mill sector of Belgium 28
Table 4.5 : Continuous 5-shift system in the German chemical industry 28
Table 4.6 : Continuous 5-shift system in the chemical industry
(additional shifts are necessary) 29
Table 4.7 : Continuous 5-shift system in the Dutch, English, French and German
chemical industry 29
Table 4.8: Continuous 5-shift system in Dutch electrical plant 30
Table 4.9: Continuous 5-shift system in the Dutch electrical sector 30
Table 4.10 (a) : Typical 5-crew shifts in the Mars factory 31
Table 4.10 (b) : 5-crew shifts in the Mars factory during summer holiday period 31
Table 4.11 : Continuous 6-shift system in a German nuclear power plant 33
Table 4.12 : Summary table of the successive system (extract from the Report of the
Walloon station) 34
Table 4.13 : System of replacements (extract from the summary of the Report of
the Flemish station) 34
Table 4.14 : Continuous 2-shift system in the Belgian petroleum industry 35
Figure 4.1 : Distribution of workers on continuous shift systems according to the
number of crews 36
Table 4.15: 5-crew system with seven consecutive nights 37
Table 4.16: Modification of 5-crew system above 38
Table 4.17 : Alternative modification of 5-crew system 38
Table 4.18: 2-2-3 with 5th week free for Days/Training/Holidays 38
Table 4.19: 2-2-3 with 9th and 10th week free for Days/Training/Holidays 39

4
Network

Giovanni Costa Tél: 39 45 807 42 96

Istituto di Medicina del Lavoro Fax: 39 45 809 86 27
Università degli Studi di Verona E-mail: [email protected]
Policlinico Borgo Roma
37134 Vérone
Italy

Charles Gadbois Tél: 33 1 4410 78 21

Ecole Pratique des Hautes Etudes Fax: 33 1 44 41 71 69
Laboratoire d’Ergonomie Physiologique
et Cognitive
41, rue Gay Lussac
75005 Paris
France

Ben Jansen Tél: 31 20 404 40 42

Atos Beleidsadvies en-onderzoek bv Fax: 31 20 404 46 76
Gelderlandplein 75d E-mail: [email protected]
1082 LV Amsterdam
The Netherlands

Peter Knauth Tél: 49 721 608 45 64

Institut für Industriebetriebslehre Fax: 49 721 75 89 09
und Industrielle Produktion (IIP) E-mail: [email protected]
Abteilung Arbeitswissenschaft
Universität Karlsruhe, Hertzstrasse 16
76187 Karlsruhe
Germany

Robert Léonard Tél: 32 50 607 120

NV Lammeken Fax: 32 50 607 120
ERGOLAM - Department of Ergonomics
Lammekenslaan 14
8300 Knokke-Heist
Belgium

Alexander Wedderburn Tél: 44 131 44 95 111

Department of Business Organization Fax: 44 131 45 13 296
School of Management Email: [email protected]
Heriot-Watt University
Edinburgh EH14 4AS
Scotland

5
Chapter 1: INTRODUCTION

Continuous shift systems - those that run 24 hours a day every day of the year
– are the most intensive form of shiftwork. They inevitably involve night
work, which is generally considered the hardest for the human body to
tolerate; and necessarily involve work at weekends, which is, for most people,
the most alienating from the social life of their community.
In many cases, too, they are used in situations where there is no choice. It
can be argued that people working in sectors where there is no choice have a
stronger commitment to shiftwork, so that they are more likely to adapt to it
(or choose to work elsewhere).
In many service situations, such as law enforcement, medical services, and
the provision of electricity, water and gas, the public expect a service that is
always there. Of course, in remote parts of the country, such as the Scottish
islands, there may not be a policeman or ambulance constantly in
attendance; but even there, the public expect the service to mobilise itself
when there is a call for it.
In several industrial processes, there is a “technological imperative”. It is not
possible to switch off a chemical plant or steel furnace without going through
a complex shut-down procedure, which is costly and therefore unthinkable in
practical terms. However, other sectors that do not have the same service or
technical necessity for shiftwork are increasingly moving to continuous
shiftwork.
In the UK, the personal banking sector has been invaded by direct services,
where the customer can carry out transactions by telephone at any time of
the day or night and on any day of the year. The appeal of this type of service
is obvious. The human consequence is that telephone lines must be manned
constantly - but not, of course, in constant numbers. Some staff have to work
at night and at weekends, but the size of workforce can be tailored to the
varying demand at different times. Modern telephone systems also allow great
versatility in their operation, so that there are possibilities for concentrating
this workforce in one place to cover a large area of the country, or even of
several countries. When the Dell computer helpline in Dublin is too busy, for
instance, calls are transferred to Texas.
Some other examples of modern telework are less transparent: if you make a
telephone booking to British Airways, your call is answered by a Briton, but
they handle the booking through a computer system run in Bombay, India,
for reasons of cost and efficiency.
In many other industries, the cost of equipment and its rapid obsolescence
has driven management towards fuller utilisation of operating time. Whether
this means continuous shift systems or some less intensive form of shiftwork
has to be a calculated decision based primarily on costs, but also by looking at
the risk of errors and accidents, space for warehouses to contain buffer stock,
and the acceptability of different times of working to a workforce that may be
hard to recruit and long to train.
One consequence of the expansion of shiftwork to more intensive forms,
interestingly, has been a greater variety of forms of shiftwork. The old “week-
about” system of one week on each shift is still remarkably common, but new
6
ways of creating continuous cover are only limited by the ingenuity of
managers, their workforces, and (of course) of legislative and collective
bargaining restraints, where these apply.
Most people assume that more intensive shiftwork is bound to create more
problems for shiftworkers and their managers. Interestingly, van der Weerd
and Jansen report that, in several projects, they found that traditional semi-
continuous shift systems are more inconvenient for the workers than the
modern full continuous systems.
This issue of BEST is therefore devoted to looking at the problems and
solutions found in continuous shift systems, so that lessons learnt in one
industry or country can be passed on to others.
Two postscripts to this introduction are worth adding.
The mathematics of shift systems
Some readers may find it helpful to be reminded that working out a shift
system is the kind of problem that can sometimes be simplified by reduction
to mathematical formulae. In BEST 7, a section from Knauth and Hornberger
(1992) explained the principles, and parts of this are relevant to continuous
systems.
1
Formula 1

Nc = OT OT = IWT
or
IWT Nc

where Nc = Number of crews

OT = Operational Time
IWT = Individual Working Time

In continuous evenly-manned systems, OT = 168 (i.e. 7x24 hours), so if a 42-

hour week is worked, OT = 168 = 4, so 4 crews are required.
IWT 42

Most collective agreements have agreed working times of less than 42 hours.
168 = 33.6, so a five-crew system averages 33.6 hours of individual working time.
5

Allowing for annual and public holidays, training time and average absence
levels, this IWT of 33.6 may in fact average out quite close to agreed working
times of 37 or 38 hours a week. The next whole number of crews is 6, and
168 = 28. This, even with allowances for holidays, training and absence, is
6
likely to average out below the limits of most collective agreements on
individual working time.
Five other ways can be used to solve the problem, and also deal with the
intermediate situations where the agreed individual hours do not match the
needs of the system:
• Four crews work paid overtime to make up the difference.
• Four crews work overtime, but are compensated by additional free days.
An increase in the number of workers is needed to achieve this in order
to cover the extra free days.
1 Formula 1 is correct only when the number of employees on each shift is the same.

7
 • Five or six crews are used, but work less than their agreed working hours.
If they do not want to work part-time, they have to work additional shifts
to make up their agreed working hours. This is common with annual
working hours agreements, where the company holds a “bank” of hours
owed by individual workers, which it can call upon if needed – without
extra payment, as the hours have already been paid for. The secret of
annual hours agreements for changing a culture of overtime is that an
organisation does not require all contracted hours to be worked, unless
they are needed. This takes away the financial incentive for working
overtime.
• A combination of 4-crew and 5-crew systems: During some months of the
year the workers work in a 5-crew system (33.6 hours/week); during other
months (e.g. holiday periods) the same workers change to a 4-crew system
(42 hours/week). Over the year, the agreed working time is achieved.
This is a commonly found solution in seasonal work; and many shift
systems change patterns in the summer holiday period in a similar way.
• The crews are split into subgroups. Intermediate problems can be solved
by using combinations of subgroups, assuming that the workforce is
168
interchangeable. So 4.5 = 37.333. 4.5 crews are achieved by splitting the
workforce into 9 half-crews, two of which work at any time. (The formula
168 = 4.5 is, of course, equivalent.)
37.3

Guidelines for a good shift system

Fourteen rules were given in BEST 3, Guidelines for shiftworkers, reprinted
in Appendix 2 of BEST 7, and are worth repeating here.
1. Minimise permanent nights
2. Minimise sequence of nights: only 2-4 night shifts in succession should
be worked
3. Avoid fast double-backs
4. Plan rotas with some weekends free
5. Avoid overlong work sequences
6. Fix shift length to task loads
7. Consider shorter night shifts
8. Rotate forward
9. Delay morning start
10. Make shift change times flexible if possible
11. Keep rotas regular
12. Allow some individual flexibility
13. Limit short-term rota changes
14. Give good notice of rotas
Guidelines 1,2,3,4,5,8 and 9 have been given particular emphasis by Knauth
(1996) in more recent writing – with a limit of 3 on the number of
2
nightshifts that should be worked in succession (point 2 above).
The following sections go fairly rapidly into a huge number and variety of
shift systems. It is hoped that these two pieces of general wisdom will help to
throw light on the picture.

2 It is well worth reading the full descriptions of the guidelines in BEST 3 to understand them, as this
gives the evidence and argument that lie behind them.

8
Chapter 2: PREVALENCE (FREQUENCY) OF
CONTINUOUS SYSTEMS
Italy
In Italy, there are no official statistics on the current prevalence of
continuous shift systems among the various work sectors. (Costa, 1995)

Belgium
In Belgium, Léonard (1995) reports that a 1977 university survey (K.U.L.,
1980) of 913 companies with more than 50 workers indicates that a shift
system was used in 37% of those companies and that in 26% of these, several
systems could be used at the same time, so that the frequency of continuous
work was 24%. These figures should be compared to more recent data
(National Institute of Statistics, 1994). That data is summarised in Table 2.1
and shows that, of 3,749,700 people, 562,341 (15%) worked shiftwork and
531,903 (14%) worked at night. The K.U.L. survey only takes into account
industrial companies, whereas the data of the National Institute of Statistics
concern all employed workers.
Out of 100 shiftworkers in 1977, the distribution of the systems is as follows :
• discontinuous : 31%
• semi-continuous : 40%
• continuous : 23%
• other systems : 6%

89% of shiftworkers were men and 11% women. This percentage is changing,
and at present 16% of shiftworkers are women. Production workers work
more often in shifts than management (38% compared to 7%, according to
Mairiaux 1995), but those figures also date from 1977.
Continuous work is found mainly in the following sectors : textile industry,
metallurgy, glass and ceramics, chemistry, metal manufacture, paper, oil
industry, electricity works, and food. It is also found in health services,
firefighting, and transport.

9
10

Table 2.1 : Population with a job: shiftwork, irregular hours, by economic sector: Belgium (1994)

Branch of economical activity

Shiftwork finances, administration,

irregular agriculture, extractive manufacturing electricity, construction trade hotels, transport property, education, domestic TOTAL
hours, work fishing industries industries gas, water restaurants business health services
at home

Shiftwork:
– usually 707 2,898 203,103 3,907 13,790 30,027 12,130 47,608 17,628 147,063 2,437 481,298
– sometimes 217 3 16,936 1,148 1,569 6,767 1,856 9,712 6,392 35,812 294 81,043
– never 107,343 40 556,570 24,828 249,004 519,327 102,109 202,678 345,916 1,053,566 17,108 3,187,360
8,311

Evening:
– usually 29,254 1,879 114,334 1,369 5,831 52,505 63,394 38,552 34,049 135,857 1,834 478,867
– sometimes 41,880 1,982 138,032 5,540 30,500 112,198 23,853 59,498 71,122 66,355 3,345 754,305
– never 37,139 7,689 524,233 22,977 228,042 391,414 28,862 161,953 264,772 834,239 15,252 2,516,572

Night:
– usually 2,090 1,413 61,601 954 1,111 12,737 17,877 25,939 5,259 61,255 671 190,907
– sometimes 38,978 1,213 59,874 4,123 6,005 16,611 19,783 40,436 16,591 135,341 2,041 340,996
– never 67,204 8,925 655,125 24,809 257,246 526,766 78,446 193,622 348,085 1,039,841 17,728 3,217,797

Source: National Institute of Statistics, Belgium (1994)

Germany
In Germany, Knauth and Hornberger (1995) report on a 1993 questionnaire
completed in West Germany (Bauer et al., 1994), which indicated that the
frequency of shiftwork had decreased from 14% (in 1989) to 12%. They
suggest that this might be due to new “shiftlike” working time arrangements,
such as staggered working time or the combination of full- and part-time
regulations. However, there was no change in the number of employees
working on Saturdays (29%) or Sundays (12%), and almost all employees
(93%) who worked on Sundays worked on Saturdays as well. There were
more male (14%) than female (9%) shiftworkers. As shown in Table 2.2, the
bigger the company, the higher the frequency of shiftwork. In the smallest
enterprises (i.e. those with less than 5 employees) the shiftwork rate is as low
as 2%; in enterprises with more than 500 employees, 23% do shiftwork.

Table 2.2 : Frequency of shiftwork by size of enterprise (in %)

Employees in companies with (%) working shiftwork
Frequency of 1-4 5 - 99 100 - 499 500 and employees
shiftwork employees employees employees more total
employees
regularly 2 6 15 23 12
seldom/never 99 94 85 77 88

Source: Bauer et al. (1994)

42% of employees work in a 2-shift system and 31% in a 3-shift system. 7% of

these work continuous shifts, and 39% work night shifts.

Netherlands
From the Netherlands, van der Weerd and Jansen (1996) report that,
although working outside the usual “nine-to-five” routine is becoming more
and more common, it is characterised not by a higher percentage of people
working in round-the-clock systems but by an increase in the variety of such
systems. One cause for this is the increasing need of trade and industry to
utilise more flexible and efficient ways of employment, related, in turn, to an
increasing social need to take into account employees’ preferences regarding
their leisure time.
In Dutch industry during the period 1982-1988, the percentage of men
working in 2- or 3-shift systems decreased by approximately 1% while the
percentage working in round-the-clock shifts increased by about 2% (FNV,
1992). About 70,000 employees are working regular continuous shifts in the
Netherlands (about 1.5% of the workforce). According to van der Weerd and
Jansen, traditional forms of shiftworking in the Netherlands are going more
and more out of fashion, and “tailor-made” solutions are becoming the rule,
rather than the exception, both in daytime work and in 24-hour systems.
“These arrangements are varying per sector, per company and per department
and even sometimes per group of employees or per individual employee. This
makes it hard to distinguish fully continuous from semi-continuous working
hours...”
The most recent Labour Force Survey for the Netherlands (CBS, 1996)
reports the following percentages of employees working shiftwork:

11
 • regular 52%
• irregular 48%
• – night/evening 15%
• – evening 17%
• – weekends (daytime) 16%
Van der Weerd and Jansen also note that “the higher demands to which
(flexible) working hour regulations are subject have been laid down in the
Arbeidstijdenwet (Working Hours Act) (ATW) in the Netherlands, which
came into force for most sectors on 1 January 1996, and in many recent
collective labour agreements about the introduction of the average 36-hour
week.” (Van der Weerd and Jansen, 1996)
France
In France, Gadbois and Dorel (1996) report that a national survey by the
French Ministry of Labour has provided a precise and highly detailed measure
of the practice of continuous shiftwork in the various economic sectors, and
describes trends over the last decade (Bloch-London, 1992). Between 1981
and 1990, the prevalence of shiftwork in France increased from 11 to 12.5%,
the greatest increase being among blue-collar workers (18.5 to 22.6%). In
1990, 21% of all shiftworkers were on continuous shiftwork. The percentages
were lower among blue-collar workers (17%) than among other employees
(40.2%) (Table 2.3).

Table 2.3 : Distribution of shiftworkers according to shift systems in France

(1990)
Discontinuous Semi-continuous Continuous
Blue-collar workers 54.7% 28.3% 17.0%
Other employees 39.0% 20.8% 40.2%
All employees 51.9% 26.9% 21.2%
Source: Bloch-London (1992)

The prevalence of continuous shiftwork varies greatly in different sectors.

Figure 2.1 indicates the proportion of shiftworkers among all employees for
each sector, and the proportion of continuous shiftworkers among all
shiftworkers. Some sectors, such as energy and water, have few shiftworkers,
but most of these work continuous shifts. Other sectors, such as rubber and
plastics, have many shiftworkers, but few work continuous shifts.
The following industrial sectors depend the most on continuous shiftwork:
– energy (petroleum, gas, electricity): 4 in 5 employees,
– glass, industrial chemistry (synthetic and artificial fibres): 2 in 3,
– steel, paper and paper products: 1 in 3,
– building materials and minerals: 1 in 5,
– food (except for meat and milk): 1 in 5 to 1 in 10,
– chemicals, textiles-clothing, printing and publishing: 1 in 5 to 1 in 10.
Outside industry, more than half of shiftworkers are on continuous shift
systems in the health and transport sectors.
A detailed analysis of shiftwork in different specific industries can be found in
Gadbois and Dorel (1996).
12
Figure 2.1 : Continuous shiftwork in different sectors in France (1990)

Minerals ferrous metals steel

Glass

Rubber & plastics

Paper & paper products

Minerals & non ferrous metals

Matériel de transport terrestre

Industrial chemistry, artificial &

synthetic fibres

Food industry (meat & milk apart)

Food retailing

Metal manufacture

Textiles, clothing

Petroleum, gas

Construction materials

Meat & milk % of shiftworkers

Printing & publishing % of workers on continuous shifts

Electrical & electronics

Consumer services

Land transports

Chemicals & pharmaceuticals

Mechanical construction

Timber & wooden furniture

Shipbuilding & aerospace

Hotels & catering

Energy & water

Food wholesaling

Footwear & leather

Services to companies

Wholesaling (food apart)

Motor vehicles & parts

0 10 20 30 40 50 60 70 80 90

13
 United Kingdom
In the UK, shiftwork questions are asked annually in the spring (March to
May) Labour Force Survey. The questions are asked consistently from year to
year, so that trends can be observed. However the main questions are not
very informative for understanding key information about what is happening
to shift systems generally.
The first question is simply: “Do you do shiftwork in your (main) job?” with
answers of “usually, sometimes, never”.
Results from this show that the percentage of employees defining themselves
as shiftworkers has risen slightly each year over the 1992-1994 period (see
Table 2.4, below). The main problem is in the ambiguity of the word “shift”,
as this means “a changing working time” to some people, but “non-daywork”
to others. Thus there is 11% of the employed population that report working
at night, but who do not describe themselves as shiftworkers.

Table 2.4 : Percentage of employees working shifts: UK

March/May 1992 March/May 1993 March/May 1994
Usually Sometimes Usually Sometimes Usually Sometimes
% % % % % %
male 8.3 2.2 8.6 2.2 8.9 2.3
female 5.6 1.0 5.9 1.2 6.2 1.2
total 13.9 3.2 14.5 3.4 15.1 3.5
M+F total 17.1% 17.9 % 18.6 %
Source: Labour Force Survey (unpublished data)

The percentage working shiftwork “usually” appears to have grown by 7% for

males and 10.7% for females between 1992 and 1994.
The second question asks shiftworkers what type of shiftwork they work. The
problem here is the answer categories, and their definitions. Table 2.5 shows
how these types of shiftwork vary for men and women; definitions of the
shiftwork types, as coded by the Labour Force Survey, follow the table.
Table 2.5 : Percentages of men and women in different types of shiftwork :
UK, 1992 vs. -1994
MALE FEMALE
TYPE OF SHIFT 1992 1994 1992 1994
3-SHIFT 12.2 11.3 4.5 4.9
CONTINENTAL 3.5 3.2 0.3 0.4
2-SHIFT 16.3 16.7 14.5 13.8
NIGHT OR DAY 10.8 9.7 4.2 3.4
SPLIT SHIFT 2.1 2.1 2.4 2.4
MORNING SHIFT 0.8 1.0 0.6 1.0
EVENING/ TWILIGHT 1.3 1.6 2.0 2.8
NIGHT SHIFT 3.6 4.7 3.6 4.1
WEEKEND SHIFT 0.3 0.4 0.4 0.4
OTHER 10.1 9.3 6.5 6.8
TOTALS: 61% 60% 39% 40%
Source: Labour Force Survey (unpublished data)
NOTE: Figures given are percentages of all those working shiftwork “usually” or “sometimes”.

14
Definitions:
Three-shift working The day is divided into three working periods - morning,
afternoon and night. This kind of shiftwork usually, but not
always, involves one or more weeks of mornings, followed by
one or more weeks of afternoons, followed by one or more
weeks of nights.
Continental shifts This is a continuous three-shift system that rotates rapidly, e.g.
three mornings, then two afternoons, then two nights. Usually
there is a break between shift changes.
Two-shift system This is normally two shifts of eight hours each, e.g. 0600-1400
with earlies and lates and 1400-2200. Shifts are usually alternated weekly or over
or double day shifts longer intervals.
Split shifts These are full shifts divided into two distinct parts with a gap
of several hours in between. Used in industries where peak
demands are met at different times of the day, e.g. catering,
passenger transport and service industries.
Morning shift If this is full-time, most commonly 0600-1400. This code is
used if the morning shift is the only shift worked or worked
part time during the morning.
Evening shift If this is full-time, most commonly 1500-2400. Also used for a
part-time shift 1700-2100 or 1800-2200. Part-time evening
shifts are usually called twilight shifts.
Night shift If this is full-time, most commonly 1800-0600, and usually
continuing after midnight. This code is used only for
permanent night work.
Weekend shift This code is used for work during Fridays, Saturdays, Sundays
(0600-1800), when there is no other work.
Other type of This code is only used when none of the above apply.
shift work

Source: Labour Force Survey (Survey User’s Guide: “Shiftwork & Weekend Working”, unpublished
instructions, March-May 1994)

It is not possible to be certain that respondents always follow the strict

definitions of the interview schedule. For example, 4-crew 12-hour shift
systems are now quite common in the UK: a respondent could call this “2-
shift” (as they only work two different kinds of shift, 12-hour days or 12-hour
nights), or “night or day” (which is strictly meant for the system of a day
shift, a gap, a night shift, another gap, and then the next day shift) or even
“continental”. The curiosity about “continental” is that it originally appears
to have been used to describe shift systems that ran through over the
weekend - now commonly called “continuous shift systems”, in the belief that
this pattern was common on the mainland of continental Europe. The first
“continental” shift systems tended to have 7 Mornings, 2 Off, 7 Evenings, 2
Off, 7 Nights, 3 Off. One of the first rapidly rotating shift systems, the 2-2-3,
was also called “fast continental” and the definers of the survey questions and
responses have picked on that as the “official” meaning of continental.
Thus the UK results are probably fairly comparable from year to year, but
must be interpreted with caution, because of the possible random assignment
of cases like the above, depending on the rapport of the interviewer and
respondent. They are certainly not easy to compare with national statistics
from other countries.
15
 Within the UK, there appear to have been substantial increases in employees
on certain shifts both between 1992-93 and 1993-94:
• Morning shifts - 42% increase;
• Evening or twilight shifts - 41% increase;
• Night shifts - 28% increase.
The figures for 3-shift working and continental shifts are shown below by
industrial sectors. Neither of these is identified with continuous shiftworking.
The most interesting point about these tables is the growth in sectors 5 and
10, combined with a decline in more traditional sectors, such as 2, 3, 4 and 8.

Table 2.6: 3-shift workers in various industrial sectors (in thousands):

UK (1992 vs. 1994)
Industry 1992 1994 %
Change
Male Female Total Male Female Total
1 0 0 0 1.3 0 1.3 NA
2 50 0.4 50.4 24 0 24 –52%
3 53 1.3 54.3 45 3.4 48.4 –11%
4 75 8 83 86 6 92 11%
5 90 10 100 97 7 104 4%
6 5 0 5 2 0 2 –60%
7 20 18 38 26 26 52 37%
8 76 8 84 79 6 85 1%
9 11 4 15 23 9 32 113%
10 128 132 260 128 163 291 12%
Source: Labour Force Survey (unpublished data)

Table 2.7 : Continental shiftworkers in various industrial sectors

(in thousands): UK (1992 vs. 1994)
Industry 1992 1994 %
Change
Male Female Total Male Female Total
2 13 0 13 10 0 10 –23%
3 43 0 43 38 2 40 –7%
4 25 2 27 24 0.4 25 –10%
5 30 2 32 35 3 38 19%
6 3 0 3 0.4 0 0.4 –87%
7 5 2 7 6 3 9 29%
8 7 0.8 8 6 0.9 7 –12%
9 1.7 0.4 2.1 3 0 3 25%
10 17 6 23 18 8 26 13%
Source: Labour Force Survey (unpublished data)

Key to industrial sectors:

1 – Agriculture, forestry and fishing
2 – Energy and water supply
3 – Minerals, ores, metals, chemicals
4 – Metal goods, engineering, vehicles
5 – Other manufacturing industries
6 – Construction
7 – Distribution, hotels and catering
8 – Transport and communications
9 – Banking, finance, business services, leasing
10 – Other services
16
Chapter 3: SECTORS WITH CONTINUOUS
SYSTEMS
Introduction
Similar industrial sectors are compiled in this chapter, because there are
probably underlying similarities in technology in Western Europe, and it may
also be more interesting to some readers to jump straight to their own sector.

Changes and Trends

Knauth and Hornberger (1995) note two main trends in the development of
continuous shiftworking in more and more companies in Germany:
• changes from slow- to fast-rotating shift schedules
• the introduction of flexibility in working time arrangements both for the
company and employees, such as flexitime, individual working time
arrangements, yearly working time or part-time work. This is partly due
to the reduction of agreed working time, as companies are being forced
to find ways to cover their operational time, which has remained
constant.
Van der Weerd and Jansen (1996) report that, while “traditional” forms of
shiftwork in the Netherlands are found mainly in the industrial sector,
irregular shifts occur mostly in the services sector, which has seen a 15%
growth in employment during the period 1987-1993. They note a two-stage
development in round-the-clock systems generally:
• “Stage 1: broadening of working hours, increase of 24-hour working
hour regulations in shift work, development from four- to five-
shift systems;
• Stage 2: switching from round-the-clock shift systems to more
irregular/flexible 24-hour systems.”
Two other developments seen from the mid-70’s have been (1) a reduction of
working time in full, continuous operations (down to 33.6 hrs/week) and, as
in Germany, (2) changes from slow- to fast-rotating shift schedules.

Textiles
In recent years the need for increasing productivity and reducing labour costs
has forced the extension of working days. In Italy, the textile sector has seen
the most extensive changes in working time organisation.
During the late 70’s and the early 80’s many companies that had operated
using classical working times based on day work (normal day or 2 shifts) on 5
days per week (from Monday to Friday) took their first step towards semi-
continuous shift systems based on three 8-hour shifts on 5 days, or on four 6-
hour shifts on 6 days per week. This extended productive time to 4,500-6,000
hours per year. In the case of 3x8 shift systems, workers operated for 40 hours
per week, whereas, in the case of 4x6 shift systems, working hours were
reduced to 36, but with the same salary. The extra four hours’ pay was
justified by the elimination of meal breaks, and by the necessity of making
one extra commuting journey to work each week.
Interesingly, the Netherlands has had some experience with 6-hour shifts in
the banking sector, which reduced the working week to about 28 hours/week;
they were enjoyed by the workers.
These 6-hour shifts were in many ways quite revolutionary, and are well
worth more study and widespread discussion.
In the late 80’s and early 90’s, many firms took a second step towards
continuous shift systems by including Sunday work, which raised the total to
17
 8,000 hours per year of plant utilisation. The shift schedules were mainly
organised with 3 or 4 shifts per day, of 8 or 6 hours respectively, but the
workers obtained a further reduction of their working hours to 34, 32 or even
30 hours per week, paid as 40.
The distribution of working days has been arranged mainly on 3 or 4 days
according to the length of shifts (6 or 8 hours) and the contracted length of
the working week. According to a recent study of 42 companies (Ravasio and
Zanzottera, 1991) the shift rotas adopted are:
• 3 days on / 1 or 2 days off: 57.1%
• 4 days on / 1 or 2 days off: 28.6%
• 5 days on / 2 or 3 days off: 7.1%
• 6 days on / 2 or 3 days off: 7.1%
There are also some cases of faster rotating shift schedules with 3 days on and
2 days off, and there is one factory that adopted 12-hour shifts, so that people
work 32 hours per week (paid as 40) on 3 days (12h-12h-8h), then have 4
days off.
In a few cases, part-time work has been introduced, integrated with full-time
work on continuous shift systems. Full-time workers are employed from
Monday to Friday, while part-time workers are employed on two 12-hour
shifts on Saturday and Sunday plus 8 hours on one normal day every two
weeks.
These shift schedules, after initial opposition connected with the
introduction of night and/or Sunday work, are now particularly appreciated
by women, because of the increased free time they can spend on their family
commitments.
From the labour market perspective, despite a 25% reduction of the
workforce suffered during the 80’s, this re-organisation of working time was
one of the main factors that allowed the Italian textile sector to maintain
competitiveness and to sustain job losses 3-4% lower than in other industrial
sectors. This sector now accounts for 1/3 of the total European workforce.
(Costa, 1995)
Léonard (1995) reports different patterns in the textile industry in Belgium: 2
shifts of 8 hours a day each, 2 shifts followed by a night shift, 3 shifts and 4
shifts (no work on Sundays), and 5 shifts. The typical company works every
day of the year. While 3 shifts are at work, two are at rest. (Table 3.1)
Table 3.1 : Shifts in the textile industry in Belgium
Cycle Mon Tue Wed Thu Fri Sat Sun
1st week M M M R R A A = 40 hours
2nd week A R R N N N R = 32 hours
3rd week R M M M M R R = 32 hours
4th week A A A A R R N = 40 hours
5th week N N N R R R R = 24 hours
=168 hours
=33.6 hours/week
Totals3 MAANR MMANR MMANR MANRR MNRRR ANRRR ANRRR
Key: M = morning, A = afternoon, N = night, R = rest

3This is, obviously, not the complete rota, as the column totals do not produce balanced manning, but
this is all that Blanpain and Kohler (eds.) show.

18
In one of the textile plants a full continuous shiftwork system has been
introduced for manual workers ensuring full production every day of the year
(including Saturdays, Sundays and public holidays, with the exception of
Christmas and New Year's Day). To that end six shifts operate in such a way
that while three shifts each perform 8 hours, the three other shifts are at rest.
Each worker works 26 weeks a year (7 days of 8-3/4 hours; one week work,
one week rest, while during the third week the worker can be called up); and
each worker has three weeks of paid holidays. Work on Saturdays is rewarded
with an extra 25% premium payment during the day (until 7.30 p.m.) and
64% at night (9.30 p.m. - 5.30 a.m.); 100% for work on Sundays, and 200%
on public holidays.

Bridging shifts
Special shifts work on Saturdays, Sundays and public holidays, while other
shifts do not work during these days. These special shifts do not work on
Saturdays and Sundays during their annual holidays, of course.
In other textile plants of the same enterprise, bridging shifts have been
introduced, assuring the continuity of the activities on 49 Saturdays, 49
Sundays and the 10 legal public holidays (the other Saturdays and Sundays
are part of the 3 weeks’ paid vacation). Two shifts, each of 12 hours, have
been introduced. Participation in the bridging shifts is voluntary.
The dates and spread of the annual holidays are determined by the joint
committee for the branch of the industry. If the joint committee fails to
decide, the decision will be taken by the works council. Since 1975, Belgian
employees generally have enjoyed four weeks’ vacation and 10 paid public
holidays. The enterprise is closed for at least two weeks during the summer,
and rest days can be added on to these holidays.

Metallurgy
In Belgium, Léonard reports that one company in the steel industry has
adopted a 7-7-7-7 system: 7M/7A/7N/7R. The only change consists in the
fact that the shifts start at 7 a.m.-3 p.m.-11 p.m. instead of the classical 6
a.m.-2 p.m.-10 p.m. During the holidays, they call upon personnel of other
departments or from outside.
Costa (1995) reports that an important steel company in Northern Italy has
recently made an agreement on the extension of continuous shift systems in
some departments of a large steel factory. This has been done by introducing
a fifth crew, combined with some solidarity contracts for young people (see
BEST 9), which allowed a reduction of weekly working hours of shiftworkers
to as little as 30 hours.
The adoption of the fifth crew led to a re-organisation of shift schedules: the
previous shift rota cycle was 20 days long, having two periods of “5 days on /
2 days off” alternating with one period of “4 days on / 1 day off”. The new
shift cycle was reduced in length to 13 days, including one period of “4 days
on / 3 day off” alternating with one period of “4 days on / 2 days off” (Table
3.2). Consequently, the shiftworkers benefit by 18 more days off per year.
(This is, strictly speaking, a semi-continuous shift system, with free
weekends.)
19
 Table 3.2 : Continuous shift system in a steel company in Northern Italy
Shift Cycle Su Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa
1 M M M M M R R R A A A A R R
2 N N N N Sn R R R M M M M R R
3 A A A A R R R N N N N Sn R R
Key: M = Morning A = Afternoon N = Night
R = Rest Sn = “Smonto Notte”

Interestingly, the first day off after night shift is not counted as a rest day. In
Italian, it is called “smonto notte”, meaning “dismount from night”. As the
first recovery day is often felt to be a wasted day because it is partly spent
asleep, this system of computation may have wider appeal.
Another important metallurgical company adopted a continuous shift system
on a 3x8 shift schedule, with four crews and fast rotation, alternating “4 days
on / 3 days off”, “3 days on / 2 days off” and “2 days on / 1 day off” shift rotas.
The compensations for shiftworkers were: a reduction of 266 hours per year;
60.4% higher pay for night hours, and a further 39.3% for weekend work; 18
minutes of overlap between the shifts, paid as overtime; and arrangement of
staggered holidays from June to September. Two 10-hour weekend shifts were
introduced for part-time workers.
Furthermore, exemption from nightwork was agreed to for women with
children under 6 years of age, unmarried women, women with old parents or
handicapped relatives requiring assistance; and women with more than 26
years’ service. It was also agreed to give particular consideration to women
whose husbands were shiftworkers.
Another company producing presses asked for an extension of working time
to Saturday and Sunday using two 10-hour shifts. The management proposed
keeping present shiftworkers on a 3x8 shift system from Monday to Friday
and starting two weekend teams working 20 hours in two days, paid as 32
hours. The unions opposed this on the ground that the people engaged in
weekend teams were not sufficiently protected. Their proposal was to extend
working time to Saturday only by increasing the number of the full-time
workers and adopting a 6x6 shift system (four shifts of 6 hours on 6 days)
with 34.5 working hours per week.
In another factory, extension to Saturday and Sunday work (but without
night work) has been made possible by making the individual’s weekly
working hours “flexible”: these ranged from 32 to 48 hours on a 6x6 shift
system. A maximum of sixteen 48-hour weeks are allowed per year, and these
must be balanced by the same number of 32-hour weeks. The eight hours
above 40 (which is contractual weekly working time) are paid as overtime.
Chemical sector
In general, in Italy continuous shift systems are organised on a basis of 244
individual working days per year (37h 20m per week on average). Most
companies in Italy adopt the 3x8 shift system with “4 days on / 2 days off” or
“2 days on /1 day off” shift rotas; normally they have four ordinary crews plus
some “floating” workers, who also alternate on to daywork. This permits the
reduction of annual working days to 240.
In some oil refineries a fifth crew has been introduced (in some cases
permanently; in others, only seasonally), thus reducing individual working
hours to 33h 36m per week on average. This is also true of most oil refineries
in the Netherlands.
20
In a large rubber company in Northern Italy, the introduction of the
continuous shift system (3x8 shifts on a 4/2 rota) has been accompanied by a
reduction of annual working days to 214.5 instead of 229 (as it was in semi-
continuous shift systems), together with an increase in salary for Saturday and
Sunday work.
In another important rubber company in Southern Italy, extension to
Saturday and Sunday work was carried out through the engagement of more
than 300 young people working only during weekends. (Costa, 1995)
The rota in Table 3.3 basically uses four consecutive 12-hours shifts, followed
by 4 days off. In every 60 days each crew has an extended break of 16 days -
see Crew B at the end of the diagram in Table 3.3.

Table 3.3: 12-hour 5-crew system, with 4 consecutive shifts: UK chemical

industry
smtw tfss mtwt fssm twtf ssmt wtfs smtw tfss mtwt fssm twtf ssmt wtfs smtw
M A C B D A E B D C E B A C E D
N B D A E B D C E B A C E D A C
O D A C B D A E B D C E B A C E
o C B D A E B D C E B A C E D A
B E E E C C C A A A D D D B B B
Source: Chemical works in the U.K.
Note: As each crew always works 4 of each shift, days of the weeks are grouped in fours, with
only one crew shown against each block of 4.
Key: M=Mornings (or Days), N=Nights, O=Days Off after Days, o=Days off after Nights,
B=extended Break

From the chemical industry in Belgium, Léonard (1995) reports briefly on

increasing variety. “The system which is often chosen is the following : a
cycle of 25 days with 5M/3R/5N/4R/5A/3R. We also find the system
7M/2R/7A/2R/7N/4R/4D/2R, or the system with fast rotation,
2M/2A/2N/4R. Another system is 5M/3R/5A/3R/5N/4R.” (Table 3.4)

Table 3.4 : Variety in shifts in the Belgian chemical industry

(a) 5M/3R/5N/4R/5A/3R

WK M T W T F S S M T W T F S S M T W T F S S M T W T F S S M T W T F S S
1 M M M M M R R R N N N N N R R R R A A A A A R R R M M M M M R R R N N
2 N N N R R R R A A A A A R R R M M M M M R R R N N N N N R R R R A A A
3 A A R R R M M M M M R R R N N N N N R R R R A A A A A R R R M M M M M
4 R R R N N N N N R R R R A A A A A R R R M M M M M R R R N N N N N R R
5 R R A A A A A R R R M M M M M R R R N N N N N R R R R A A A A A R R R
Total of each column: MANRR

(b) 7M/2R/7A/2R/7N/4R/4D/2R
WK M T W T F S S M T W T F S S M T W T F S S M T W T F S S M T W T F S S
1 M M M M M M M R R A A A A A A A R R N N N N N N N R R R R D D D D R R
2 R D D D D R R M M M M M M M R R A A A A A A A R R N N N N N N N R R R
3 N N N N R R R R D D D D R R M M M M M M M R R A A A A A A A R R N N N
4 A A R R N N N N N N N R R R R D D D D R R M M M M M M M R R A A A A A
5 R R A A A A A A A R R N N N N N N N R R R R D D D D R R M M M M M M M
Total of each column: MANRR or MANDR

21
 (c) 2M/2A/2N/4R

WK M T W T F S S M T W T F S S
1 M M A A N N R R R R M M A A
2 N N R R R R M M A A N N R R
3 R R M M A A N N R R R R M M
4 A A N N R R R R M M A A N N
5 R R R R M M A A N N R R R R
Total of each column: MANRR

(d) 5M/3R/5A/3R/5N/4R

WK M T W T F S S M T W T F S S M T W T F S S M T W T F S S M T W T F S S
1 M M M M M R R R A A A A A R R R N N N N N R R R R M M M M M R R R A A
2 A A A R R R N N N N N R R R R M M M M M R R R A A A A A R R R N N N N
3 N R R R R M M M M M R R R A A A A A R R R N N N N N R R R R M M M M M
4 R R R A A A A A R R R N N N N N R R R R M M M M M R R R A A A A A R R
5 R N N N N N R R R R M M M M M R R R A A A A A R R R N N N N N R R R R
Total of each column: MANRR

Over the last five years Knauth and Hornberger’s research team has tracked
changes in shift schedules in 29 production units in various branches of
German industry, most of which changed to (new) continuous shift systems.
One group of shiftworkers in a chemical company changed from a continuous
4-shift to a 5-shift system, because a reduction in the agreed working time
(from 39 to 37.5 hours per week) necessitated a new system in order to
manage the planning of these employees’ working time. The company used
this opportunity to introduce more ergonomically favourable shift schedules,
all of which were approved by the shiftworkers’ vote. The additional shifts
that must be worked to achieve contracted weekly working time can be used
in a flexible manner, in order to adjust working time volume to the
fluctuation of work volume.

Electricity-generating sector
The continuous shift systems adopted in power stations in Italy are based on
six crews, with fairly rapid rotation, and completing their full cycle in six
weeks (Table 3.5) This replaced a 5-week 5-crew cycle, and is generally more
popular.
It is worth noting that there are only 7 night shifts in each six week cycle (3
consecutive nights once, and 2 consecutive nights twice). There are 12 Rest
days in each 6 weeks, including one weekend with both Saturday and Sunday
off, and a total of 3 of the Saturdays free and 3 of the Sundays free. There are
also 9 Day shifts in each cycle, in one case from Monday to Friday, and in the
other from Tuesday to Friday.
The previous five-crew rota (Table 3.6) had 7 night shifts in each five weeks
(4 consecutive nights once and 3 consecutive nights once); 10 Rest Days in 5
weeks and no weekends with both Saturday and Sunday off; and 4 Day shifts
in each cycle, from Tuesday to Friday.
Compared with the old rota, the new rota also gives
• better opportunities for planning holidays without incurring overtime, and
• longer periods of Rest days, which is better for those commuting a long
distance to work.
22
Table 3.5 : Six-crew shift rota in an electric plant in Italy

Week Mo Tu We Th Fr Sa Su
1 R D D D D R N
2 N A A R R A A
3 A R M M M M R
4 R N N A A R M
5 M M R N N N R
6 D D D D D R R
Totals MANDRR MANDDR MANDDR MANDDR MANDDR MANRRR MANRRR

Table 3.6 : Former five-crew shift rota in an electric plant in Italy

Week Mo Tu We Th Fr Sa Su
1 R A A A A R N
2 N N N R R A A
3 A R M M M M R
4 R D D D D R M
5 M M R N N N R
Totals MANRR MANDR MANDR MANDR MANDR MANRR MANRR

Key: D = Daywork (0800-1230; l330-1730)

M = Morning shift (0800-1600)
A = Afternoon shift (1600-2400)
N = Night shift (2400-0800)
R = Rest day

Some general information about other work sectors

In the graphics sector in Italy, workers involved in continuous shift systems
have a reduction of working hours from 40 to 36 hours per week, plus 4 extra
days off per year.
In the paper sector in Italy, the mean working time is 37h 20m per week with
“2 days on / 1 day off”, “4 days on / 2 days off” or “6 days on / 3 days off” shift
rotas.
In the broadcasting sector in Italy, shiftworkers have a contractual working
time of 35 hours per week in state companies, and of 38.5 in private ones.
In the food sector in Italy, the national work contract foresees a reduction of
84 hours per year for continuous shiftworkers, of 76 hours for 2-shift workers
and of 80 hours for 3-shift workers on semi-continuous shift systems. (Costa,
1995)
In the hospital sector in Belgium, Léonard has recently carried out a survey
for the ONEM (National Service for Employment and Labour). His team
found 142 different schedules for 356 persons questioned.
He reports the following distribution:
• 1 shift : days: (~ 8 hours) 20 schedules for 14% of the personnel. Mainly
reserved for the heads of department or superiors. 5 days worked on
average and 2 days rest.
• 1 dayshift of 12 hours. 12 different schedules for 8% of the personnel : 3
to 4 days worked for ~ 3 days rest. Employees can work a full week and
accumulate rest days.
23
 • 2 shifts: morning-afternoon. 68 schedules, 48% of the personnel. Various
systems coexist of 5, 7, 10, 12, 15 and even 18 days of work. In this latter
case, the work period is followed by a week of rest.
• 3 shifts : rotation M-A-N. 33 schedules for 23% of the personnel. Often
10 days work and/or 7 consecutive nights.
• 1 permanent fixed night shift : 9 schedules for 6% of the personnel.
Generally 7 nights followed by 7 days’ rest. Part-timers work one week of
4 nights and one of 3 nights per month.

Transport
Another sector affected by continuous duty is public transport: in Belgium, it
is noted that 44% of employees in this sector work some nights. 27% of the
personnel of the railway company work continuous shifts. (Léonard and
Vandeville, 1994)

24
Chapter 4: WAYS OF MANNING CONTINUOUS
SYSTEMS
Introduction
The variety of ways of manning continuous shift systems, outlined in chapter
1, are shown in detail in practical examples from companies in this chapter.

Four crews : rapid or slow rotating, or fixed

In a packaging company in Belgium, Léonard (1995) reports that the
extrusion department, on continuous working shifts, changed from a 4-crew,
long cycle discontinuous system of 5M/2R/5A/2R/5N/8R to a system of
2M/2A/2N/4R to the great satisfaction of the workers. The duration of work
is 37 hours. There are 9 days off at the end of December. The shift patterns
are disrupted from 1 July to 31 August so that personnel can take 20 days of
holidays. During these summer holidays, a sixth shift is created, formed by
three persons coming from the five other shifts of the extrusion department
and other people from outside or other departments (4M/2R/4N/1R/4A/1R).
(Table 4.1)

Table 4.1 : Changes in shift patterns in a Belgian packaging company

a) Before: Semi-continuous -- 5M/2R/5A/2R/5N/8R
Week Mon Tue Wed Thu Fri Sat Sun
1 M M M M M R R
2 A A A A A R R
3 N N N N N R R
4 R R R R R R
Totals: MANR MANR MANR MANR MANR

b) After: 2M/2A/2N/4R
Week Mon Tue Wed Thu Fri Sat Sun
1 M M A A N N R
2 R R R M M A A
3 N N R R R R M
4 M A A N N R R
5 R R M M A A N
6 N R R R R M M
7 A A N N R R R
8 R M M A A N N
9 R R R R M M A
10 A N N R R R R
Totals 2M2A2N4R 2M2A2N4R 2M2A2N4R 2M2A2N4R 2M2A2N4R 2M2A2N4R 2M2A2N4R

25
 c) After: Holidays -- 4M/2R/4N/1R/4A/1R
Week Mon Tue Wed Thu Fri Sat Sun
1 M M M M R R N
2 N N N R A A A
3 A R M M M M R
4 R N N N N R A
5 A A A R M M M
6 M R R N N N N
7 R A A A A R M
8 M M M R R N N
9 N N R A A A A
10 R M M M M R R
11 N N N N R A A
12 A A R M M M M
13 R R N N N N R
14 A A A A R M M
15 M M R R N N N
16 N R A A A A R
Totals: 4M4A4N4R 4M4A4N4R 4M4A4N4R 4M4A4N4R 4M4A4N4R 4M4A4N4R 4M4A4N4R

Knauth and Hornberger (1995) note that, on a continuous shift system, four
crews would have to work an average of 42 hours per week to cover the 168
hours in a week. However, as the agreed weekly working time in most sectors
and countries is less than 42 hours per week, overtime has to be paid or
compensation has to be given in the form of additional days off. In a 4-shift
system, if overtime is not paid, every worker must be given an additional day
off almost every week, and this generally proves too difficult for the
supervisor or crew to handle. Therefore, the shorter the agreed weekly
working time, the stronger the pressure to change from four shifts to other
types of shift systems.
Tables 4.2 and 4.3 demonstrate two examples of continuous 4-shift systems.
The shift system in Table 4.2 has 8-hour shifts from Monday to Saturday and
12-hour shifts on Sunday to obtain one additional free Sunday. The forward
rotation of the shifts (M/E/N) is favourable but, from an ergonomic point of
view, the four consecutive night shifts are not. The advantage of the system
in Table 4.3 is the regular blocks of free days, in particular the two long free
weekends. However, more than two 12-hour shifts in a row are not
recommended if the work involves heavy physical or difficult mental tasks
(Knauth, 1993).

Table 4.2 : Continuous 4-shift system in the German chemical industry

Week Mo Tu We Th Fr Sa Su

1 M1 M1 A A A A

2 N1 N1 N1 M1 M2

3 A A N1 N1 N1 N2

4 M1 M1 M1

= day off N1 = night shift (22.00 - 06.00)

M1 = morning shift (06.00 - 14.00) N2 = night shift (18.00 - 06.00)
M2 = morning shift (06.00 - 18.00) A = afternoon shift (14.00 - 22.00)

26
Table 4.3 : Continuous 4-shift system in a Swiss oil refinery

Week Mo Tu We Th Fr Sa Su

1 D D N N N

2 D D

3 N N D D D

4 N N

= day off
D = day shift (06.30 - 18.30)
N = night shift (18.30 - 06.30)
In the German chemical industry, one company uses a 4-shift system which is
composed of the following regular subcycles: one 12-hour day shift, one 12-
hour night shift, followed by two days off, and so on.
The above rotas assume that constant manning is necessary. But van de
Weerd and Jansen (1996) remark on the importance of remembering the
following points when scheduling shifts for continuous systems:
• If night shifts are inevitable, one should aim at keeping the night staffing
as small as possible.
• The following universal rules apply:
• the work supply determines the staffing needs, but:
• the work supply should be directed towards convenient hours as much
as possible.
As Gadbois and Dorel (1996) note, such scheduling depends on the degree of
specialisation of the tasks being performed. Scheduling constraints are much
greater in sectors requiring a high degree of specialisation, where a suitable
number of operators qualified for each task are required in the crew at all
times (e.g. nuclear power stations). In contrast, there is much more flexibility
if there are a large number of employees with the same qualifications and a
number of positions requiring essentially similar skills (e.g. hospital nursing
departments).

Five crews
In Belgium, Léonard (1995) reports on a move from four crews to five in the
paper mill sector. A four-shift system (7M/2R/7A/2R/7N/3R) was replaced by
a continuous five-shift system for 1/3 of the personnel on one of the three
production lines (5M/3R/5A/3R/5N/4R) (Table 4.4). The working week was
reduced by two days (from 7 consecutive shifts to 5) and the rest periods were
lengthened by one day. The duration of the cycle is 25 days. After 7 cycles
(175 days), the cycle starts again on the same day. 105 days x 8 hours = 840
hours : over 25 weeks = 33.6 hours/week. To bring the total up to 36 hours,
staff will have to work 13 additional days, determined in advance and
distributed over the whole year. The company closes five times per year for
two days for holidays, and there is a shut-down of five days for Christmas.
The other days are taken between June and September, when the company
employs external personnel.
27
 Table 4.4 : Change from 4- to 5-shift system in the paper mill sector of Belgium
a) Before:
7M/2R/7A/2R/7N/3R
Week Mo Tu We Th Fr Sa Su
1 M M M M M M M
2 R R A A A A A
3 A A R R N N N
4 N N N N R R R
Total MANR MANR MANR MANR MANR MANR MANR

b) After:
5M/3R/5A/3R/5N/4R
WK M T W T F S S M T W T F S S M T W T F S S M T W T F S S M T W T F S S
1 M M M M M R R R A A A A A R R R N N N N N R R R R M M M M M R R R A A
2 A A A R R R N N N N N R R R R M M M M M R R R A A A A A R R R N N N N
3 N R R R R M M M M M R R R A A A A A R R R N N N N N R R R R M M M M M
4 R R R A A A A A R R R N N N N N R R R R M M M M M R R R A A A A A R R
5 R N N N N N R R R R M M M M M R R R A A A A A R R R N N N N N R R R R
Total of each column: MANRR

In West Germany, a five-crew system changes the scenario to fewer working

hours per week than generally agreed upon (168 hours/5 = 33.6 hours). Thus,
additional days have to be worked on days which are characterised in the
basic shift system as “free” days. Although in four- and five-crew shift systems
the agreed working time may be identical, from a psychological point of view,
workers would rather have additional days off than additional working days.
A few progressive companies offer part-time contracts in connection with the
introduction of a 5-shift system. A German company in the chemical
industry offers the choice between 33.6, 35.0 and 37.5 hours per week. If, for
instance, a shiftworker chooses 33.6 hours per week (with a correspondingly
reduced income), he does not have to work any additional shifts. About 20%
of the continuous shiftworkers have chosen to work less than 37.5 hours per
week. Younger and older shiftworkers, in particular, are interested in reduced
weekly working hours.
A simple way of changing from a four- to a five-shift system is to add a free
week to the old four-shift system, which has been done in Table 4.5.
However, during the “free” week, additional shifts have to be worked to reach
the agreed weekly working time. The lesser this agreed weekly working time
is, the fewer additional shifts have to be worked.

Table 4.5 : Continuous 5-shift system in the German chemical industry

Week Mo Tu We Th Fr Sa Su

1 M M A A N N N

2 M M A A A

3 N N M M M

4 A A N N

Total MAN MAN MAN MAN MAN MAN MAN

cover:

= day off A = afternoon shift

N = night shift M = morning shift

28
If shiftworkers are allowed to participate in the design of a new shift system,
the acceptance of this new rota is much higher. The five-shift system in Table
4.6 was designed by shiftworkers after receiving intensive information on
ergonomic recommendations. As with every shift system, it has also negative
aspects; e.g. it looks very irregular and has no forward rotation. The positive
aspects are short blocks of morning, evening and night shifts, a maximum of
five working days in a row and two longer free weekends.

Table 4.6 : Continuous 5-shift system in the chemical industry (additional

shifts are necessary)
Week Mo Tu We Th Fr Sa Su

1 M1 M1 N1 N1

2 A A A M1 M1 M2

3 N1 N1

4 M1 M1 N1 N1 N2

5 A A A

Total MAN MAN MAN MAN MAN MAN M2N2

cover:

= day off N1 = night shift (8 hours)

M1 = morning shift (8 hours) N2 = night shift (12 hours)
M2 = morning shift (12 hours) A = afternoon shift

The five-shift system in Table 4.7 has a very regular pattern, forward rotation,
short blocks of morning, evening and night shifts and some longer free
weekends. If an additional morning is added before the first morning shifts in
the plan, or if an additional night shift is added after the second night shifts,
the block of working days increases to seven days. As mentioned above, some
companies offer part-time work of 33.6 or 35.0 hours per week, which reduces
the amount of necessary additional shifts to zero or to one in five weeks.

Table 4.7 : Continuous 5-shift system in the Dutch, English, French and
German chemical industry

Week Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su

1+2 M M A A N N M M A A

3+4 N N M M A A N N

5+6 M M A A N N M M

7+8 A A N N M M A A N N

9+10 M M A A N N

Totals:MAN MAN MAN MAN MAN MAN MAN MAN MAN MAN MAN MAN MAN MAN

= day off
M = morning shift
A = afternoon shift
N = night shift

29
 The continuous 5-shift system shown in Table 4.8 has two features of special
interest: (1) to compensate employees on the morning shift for the
inconvenience of its heavy workload, the shift is shortened to 7 hours; and
(2) the employees work a 12-hour shift on Sundays to gain an extra Sunday
off.

Table 4.8: Continuous 5-shift system in Dutch electrical plant

WEEK Mo Tu We Th Fr Sa Sun
1 M M M M M
2 N N N N
3 A A A M M12
4 N N N12
5 A A A
Total MAN MAN MAN MAN MAN MAN M12N12

Table 4.9 shows another continuous system, the core of which is a regular 5-
shift system, worked in a cycle of 15 weeks (3x5). Each crew totals a number
of (K-L) shiftworkers. The L workers have a daytime schedule to do a project
for 15 weeks; after this period they return to the regular shifts, and another
group of shiftworkers starts or continues a project (multicraft system).

Table 4.9: Continuous 5-shift system in the Dutch electrical sector

Week Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su
1 M M M – – A A D D D D D – –
2 A A – – N N N D D D D D – –
3 – – – M M M M D D D D D – –
4 – – A A A – – D D D D D – –
5 N N N N – – – D D D D D – –
Total MAN MAN MAN MAN MAN MAN MAN 5D 5D 5D 5D 5D – –

Table 4.10 (a and b) below demonstrates how the summer holiday season is
handled by a 5-crew shift system working in the Mars factory in Slough,
England. During non-holiday months, each crew works 3M/3A/3N/6R.
During June, July and August, each has a block of 17 days in which to take
their holidays. The remaining crews work a shorter cycle which contains only
2 or 3 rest days instead of 6 and keeps alternate Saturday/Sunday blocks
together. These coincide with the end of one group’s holiday and the start of
the next group’s.

30
Table 4.10 (a) : Typical 5-crew shifts in the Mars factory

Week M Tu We Th F S S

1 M M M A A A N

2 N N R R R R R

3 R M M M A A A

4 N N N R R R R

5 R R M M M A A

6 A N N N R R R

7 R R R M M M A

8 A A N N N R R

9 R R R R M M M

10 A A A N N N R

11 R R R R R M M

12 M A A A N N N

13 R R R R R R M

14 M M A A A N N

15 N R R R R R R
3M3A3N6R 3M3A3N6R 3M3A3N6R 3M3A3N6R 3M3A3N6R 3M3A3N6R 3M3A3N6R
Totals

This is essentially a 15-day cycle, made up of 3 mornings, 3 afternoons, 3

nights, and 6 rest days. So this moves on one day every fortnight, and the
block of 3 mornings changes from starting on Monday in week 1 to starting
on Tuesday in week 3, achieving a full cycle at the start of week 16.

Table 4.10 (b): 5-crew shifts in the Mars factory during summer holiday period

Week M Tu We Th F S S

1 M M M A A A N

2 N N R R M M M

3 A A A N N N R

4 R R M M A A A

5 N N N R R R R

6 R R R R R R R

7 R R R R R R M

8 M M A A N N N

9 R R R M M M A

10 A A N N R R R
2M2A2N4R 2M2A2N4R 2M2A2N4R 2M2A2N4R 2M2A2N4R 2M2A2N4R 2M2A2N4R
Totals

In the holiday period, each shift in turn gets a break starting with their
normal rest days but running on for 17 days instead of 6, and to accommodate
this, the other shifts change to blocks of 3 (and occasionally 2) rest days,
instead of the block of 6. The big blocks of 15 (to get round from starting
mornings to starting mornings again) change to a block of 12 followed by a
31
 block of 11, thus making a bigger block of 23. This is done by making the
Wednesday-Thursday of every second week into a run of 2 on the same shift,
rather than a run of 3. The system does not run a full cycle until 10 weeks
have elapsed, by which time each of the five shifts has had its holiday break,
by adding 8 days in the middle of their normal rest days.
This is an elegant pattern, whose full symmetry with variations shows up
more clearly below when the interlocking patterns of all 5 shift crews are
shown:

Typical working month, e.g. April 1996

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
M
A
N
O
O

Holiday schedule: June 1996

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
M
A
N
O
O

Holiday schedule: July 1996

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
M
A
N
O
O

Holiday schedule: August 1996

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
M
A
N
O
O

32
Six crews
Knauth and Hornberger (1995) note that, in cases where there is a great need
for regular training (e.g. nuclear power stations), it is useful to build training
days into the shift system. The 168 hours in a week are divided by six, so that
on average each crew works only 28 hours per week. The difference between
the agreed weekly working time may be covered by additional training days.
Table 4.11 shows a continuous 6-shift system, which is used in a German
nuclear power plant.
Table 4.11 : Continuous 6-shift system in a German nuclear power plant

Week Mo Tu We Th Fr Sa Su

1 M1 M1 A A N1 N1

2 D D D D D

3 N1 N1 M1 M2

4 A A N1 N1

5 D D D A A N2

6 M1 M1 M1

Totals: MAND MAND MANDD MANDD MAND MAN M2N2

RR R R R RR RRR RRRR

= day off
M1 = morning shift (8 hours)
M2 = morning shift (12 hours)
N1 = night shift (8 hours)
N2 = night shift (12 hours)
A = afternoon shift (8 hours)
D = day shift (training: 8 hours)

A power station in Belgium, alongside a reduction of working time, went

from 4 to 5 and finally to the current number of 6 crews. This has been
achieved by adding a week of day shift to the block of five-week rosters. Table
4.12 demonstrates what has been realised by these changes, while Table 4.13
gives an idea of the replacements.

33
 Table 4.12: Summary table of the successive system (extract from the Report
of the Walloon station)

MTWTFSS MTWTFSS MTWTFSS MTWTFSS

48 h M M M M M M° AAAAAA N N N N N N N° DDD
4 shifts 46 h M M M M M M° AAAAAA N N N N N N N° DD
44 h M M M M M M° AAAAAA N N N N N N N° D

MTWTFSS MTWTFSS MTWTFSS MTWTFSS

5 shifts 42 h M M M M M M° AAAAAA N N N N N N N° D D D D D° d
40 h M M M M M M° AAAAAA N N N N N N N° D D D D°

MTWTFSS MTWTFSS MTWTFSS MTWTFSS MTWTFSS

40 h M M M M M M° AAAAAA N N N N N N N° DDDD D D D D D°
6 shifts 38-3/4 h M M M M M M° AAAAAA N N N N N N N° DDD D D D D D°
38 h M M M M M M° AAAAAA N N N N N N N° DDD D D D D D°

Source: Haumont and Gobert (1979)

Key:
M = morning 1: 7 am - 2 pm
M° = morning 2: 7 am - 6 pm
A = afternoon: 2 pm - 10 pm
N = night 1: 10 pm - 7 am
N° = night 2: 6 pm - 7 am
D = day 1: 7.15 am - 3.45 pm
D° = day 2 : 7.15 am - 3 pm
d = work one Saturday morning out of two
ö = the days of rest are left blank

Table 4.13: System of replacements (extract from the summary of the Report of
the Flemish station)
S S M T W T F S S M T W T F S S M T W T F S S M T WT F S S M T WT F S S M T WT F
A 61 62 63 64 65 66 67 R – 14a 14b 14c 14d 14e 141 142 223 224 225 226 227 22a 22b R R R R R R – R8 8 8 8 R – 8 8 8 8 8

R1 –2 83 84 8 86 87 6 6 6 6 6 6 6
5
B R – 14 14 14 14 14 14 14 22 22 22 22 22 22 22 R R R R R R -- R 8 8 8 8

C R – R 8 8 8 8 R – 8a 8b 8c 8d 8e 6 6 6 6 6 6 6 R – 14 14 14 14 14 14 14 22 22 22 22 22 22 22 R R R R R

R –2 83 84 85 86 8
1
D 22 22 R R R R R R – R 8 8 8 8 6 6 6 6 6 6 6 R – 14 14 14 14 14 14 14 22 22 22 22 22

E 14 14 22 22 22 22 22 22 22 R R R R R R – R 8 8 8 87 Ra –b 8 8 8 8 8 6 66 6 6 6 6 R – 14 14 14 14 14

F R – 14 14 14 14 14 14 14 22 22 22 22 22 22 22 R R R R R R – R 8 8 8 8 R – 8 8 8 8 8 6 6 6 6 6 6 6

Source: Haumont and Gobert (1979)

Key:
6: morning shift - 6 am to 2 pm 8: day shift - 7.55 am to 4.10 pm
14: afternoon shift - 2 pm to 10 pm - 7.55 am to 3.25 pm
22: night shift - 10 pm to 6 am R: rest
superscript : Superscripted worker replaces correspondingly subscripted worker in row above.
subscript/

34
Twelve-hour shifts
A system of 2 shifts of 12 hours succeeding each other from 6 a.m. to 6 p.m.
and from 6 p.m. to 6 a.m., with sequences of 2 or 3 days work interspersed
with 2 or 3 days rest, is reported by Léonard (1995):

Table 4.14 : Continuous 2-shift system in the Belgian petroleum industry

Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su
Team 1 – D D – – N N N – – D D D D
Team 2 N – – D D – – – N N – – D D
Team 3 – N N – – D D D – – N N – –
Team 4 D – – N N – – – D D – – N N
Totals: DN DN DN DN DN DN DN DN DN DN DN DN DDN DDN

and for the two following weeks:

Team 1 – N N – – D D D – – N N – –
Team 2 D – – N N – – – D D – – N N
Team 3 – D D – – N N N – – D D – –
Team 4 N – – D D – – – N N – – D D
Totals: DN DN DN DN DN DN DN DN DN DN DN DN DN DN

Key: D = Day N = Night

Léonard reminds his readers that there have been some negative reports
about 12-hour shifts (see also BEST 10, Compressed Working Time).
Changing from 8 to 12 hours can have a negative influence on performance
and alertness, as in the example reported by Rosa (1993) of the night shift in
a public natural gas utility company.
Gadbois and Dorel report that the number of crews on continuous shiftwork
varies considerably with the work sector (Figure 4.1), but the 3-crew system is
used by half or more workers in many sectors.

35
 Figure 4.1: Distribution of workers on continuous shift systems according to
the number of crews

19

17

34

15

18

31

20

11
Professional sectors

22

21

23

16

13

26

12

33

10

14

25

30

27

29

0% 20% 40% 60% 80% 100%

Others

6 Crews & more

5 Crews

4 Crews

3 Crews

[Key to sectors on following page]

36
Key to Sectors, Figure 4.1:
2 - Meat and milk 17 -Shipbuilding and aerospace
3 - Food industry (meat and milk apart) 18 -Textiles, clothing
5 - Petroleum, gas 19 -Footwear and leather
6 - Energy and water 20 -Timber and wooden furniture
7 - Minerals, ferrous metals, steel 21 -Paper and paper products
8 - Minerals and non-ferrous metals 22 -Printing and publishing
9 - Construction materials 23 -Rubber and plastics
10 -Glass 25 -Food wholesaling

11 -Industrial chemistry, artificial and 26 -Wholesaling (food apart)

synthetic fibres 27 -Food retailing
12 -Chemicals and pharmaceuticals 29 -Motor vehicles and parts
13 -Metal manufacture 30 -Hotels and catering
14 -Mechanical construction 31 -Land transports
15 -Electrical and electronics 33 -Services to companies
16 -Land transport materials 34 -Consumer services

Poor systems
Some companies change to five-crew systems with little understanding of
how shift systems can be best constructed. One recent example, on which a
shiftworker made contact, is shown in Table 4.15.

Table 4.15: 5-crew system with seven consecutive nights

Week S M T W T F S
1 R A A A R M M
2 M M R R A A A
3 A R M M R N N
4 N N N N N R R
5 R D D D M D R
Totals MANRR MANDR MANDR MANDR MANRR MANDR MANRR

Key: M: Morning/ A: Afternoon/ N: Night/ R: Rest Day/ D: Normal Day

The most serious criticism of this rota is the sequence of 7 consecutive night
shifts in weeks 3 and 4. In the past, BEST had recommended a maximum of
4, and some authorities (e.g. Knauth 1995) now recommend a maximum of 3.
The reason for this is that most people do not sleep for long enough during
day-sleeps, and so accumulate a sleep debt that increases the longer the
sequence goes on. There is an argument the other way, that people get used
to night shift if there are more consecutive ones, but measurements on real
shiftworkers tend to show that this is very slow, and takes 21 nights.
It is possible to modify this rota to avoid this problem, as shown in Table 4.16
below. The shifts that have been exchanged are in italics.

37
 Table 4.16: Modification of 5-crew system above

Week S M T W T F S
1 R A A A R M M
2 M M D R A A A
3 A R M M R N N
4 R D N N N R R
5 N N R D M D R
Totals MANRR MANDR MANDR MANDR MANRR MANDR MANRR

The rota in 4.16 is simply a “tweaked” version of 4.15, and gives the same
totals for each day of the week. Night shifts are now in blocks of 2, 3 and 2.
Of course the Day shifts may be in the same week in order to make it easier
to fit in holiday weeks, so that this may not be the perfect answer. It is also a
very irregular rota, which may be because there are some other purposes that
it fulfils, or possibly because it has simply evolved from earlier rotas.
Another alternative is shown below in Table 4.17.

Table 4.17 : Alternative modification of 5-crew system

Week Mo Tu We Th Fr Sa Su
1 D M M A A N N
2 R R D M M A A
3 N N R R D M M
4 M D A N N R R
5 A A N R R R R

This has the following advantages and disadvantages:

Advantages: Disadvantages:
forward rotation maximum length of working period 8 days
maximum length of nightshift
period 2 days
no unfavourable shift sequences
improved freetime on weekends
improved clarity
Another alternative, shown below as Table 4.18, would be to use a variation
of the 2-2-3, which started off as a 4-crew rota, with a 5th week added that
can either be used for normal Dayshift, for training or for holidays (marked
“dr”).
Table 4.18: 2-2-3 with 5th week free for Days/Training/Holidays

Week M T W T F S S
1 M M A A N N N
2 R R M M A A A
3 N N R R M M M
4 A A N N R R R
5 dr dr dr dr dr dr dr
Totals MANRdr MANRdr MANRdr MANRdr MANRdr MANRdr MANRdr

38
If holidays are important for the rota, it is possible to double the first 4 weeks
and then have 2 of the 5th line, giving 2 weeks free for one person’s holidays
every 10 weeks, but otherwise used for day shift (or covering absence), shown
in Table 4.19 following:

Table 4.19: 2-2-3 with 9th and 10th week free for Days/Training/Holidays

Week M T W T F S S
1 M M A A N N N
2 R R M M A A A
3 N N R R M M M
4 A A N N R R R
5 M M A A N N N
6 R R M M A A A
7 N N R R M M M
8 A A N N R R R
9 dr dr dr dr dr dr dr
10 dr dr dr dr dr dr dr
Totals 2xMANRdr 2xMANRdr 2xMANRdr 2xMANRdr 2xMANRdr 2xMANRdr 2xMANRdr

The advantage of a regular rota like this is that it is much more systematic.
Although it may look confusing at first, once the underlying pattern is
understood, it is really very simple.

39
Chapter 5: UNEVENLY MANNED
CONTINUOUS SYSTEMS
Many long-standing shift systems require continuous cover, but not evenly
manned continuous cover. This is obviously true of police, medical, and
transport systems, some of which have operated uneven continuous cover for
hundreds of years. They tend, on the whole, to use “traditional” systems,
developed over time, which have become strongly entrenched as custom and
practice. Relatively recent arrivals in the world of unevenly manned
continuous systems, often associated with the development and use of
modern technology, are less constrained by tradition, and are in some ways
more interesting.
Modern technology makes it possible for telephone calls to be transferred
locally, nationally and even globally, and to incorporate pictures as well
where necessary. So one answer to night shift manning of a technology
control centre is to divert the system to another country where it is not
night. This makes it possible to reduce to a minimum the number of people
working at night, which is one of the basic principles for the design of shift
systems.
It is difficult to give concrete examples of systems like this, but a world-wide
news service is believed to transfer overall control so that the manager in
charge is always a day worker. A system for the surveillance of supermarkets
in the US is reported to be controlled from South Africa at night. It is
thought that many world systems, such as military and air traffic control, use
similar transfers of responsibility.
In contrast, there are sometimes advantages in retaining control within one
country, where there is only a national organisation. Most people lose credit
cards during their waking hours, so that the service to receive reports of this,
and take appropriate action to stop fraudulent use of a lost credit card, is
mainly day-active. However, credit card holders from one country travel the
world for business and pleasure, so small numbers may need to report a loss
during daytime in Australia. In this case, it makes sense to maintain a
minimum staffing during night shift.

Telephone call centres

So one area where the flow of work is not even, but is to some extent
predictable, and where continuous customer service is required, is on
“telephone help lines”, which have expanded considerably in the UK in the
past five years.
The National West Bank Action Line customer telephone centre is based at
Harrogate in Yorkshire, but receives calls from all over the United Kingdom
for the price of a local call. Past trends are used to predict the volume of calls,
and are reviewed on a monthly basis. The call centre is set up so that voice-
activated routing answers most customer queries, but a human help line is
available all the time.
This is achieved by a combination of full-time and part-time staff, who are
recruited for their flexibility. The full-time Monday to Friday day staff work
in teams of about ten for three weeks of 0900-1700, one week of 1000-1800
and one week of 1200-2000. Night shifts, after 2000, are covered by
volunteers, most of whom prefer to work permanent night shifts, although
40
they may also choose to alternate this with some day shifts. Part-timers are
considered ideal for covering the “twilight” shifts from 1800-2200, and also
for Saturdays and Sundays. Very little is paid as an “unsocial hours” night or
weekend premium, so that by and large, people choose the shifts that suit
them best anyway. When trends show that the work load has changed,
management requirements are put to groups of staff, who are free to discuss
and decide the way they want to staff the system.
Another telephone banking and credit card centre has some full-time staff,
mainly at the management level. Three shift controllers work one week from
0800-1600, one week from 1600-2400 (including Saturday and Sunday), and
the third week as a combination of Rest Days, to compensate for working the
weekend, and 0800-1600. Three shift administrators work a similar morning
and evening rota. Six trainers work a rota working afternoon shifts one week
in six, and working one in six weekends. Team leaders work 35 weeks of
mornings and 15 weeks of evenings. Night shift has to be covered, so that
credit card holders worldwide (and bank customers, too) can always access
somebody.
The core of the night shift is staffed by six people, who work in 3 teams of 2.
However it is only during the period from 0100 to 0600 that staffing falls to
two, as some of the part-time staff work on shifts that end at 0100.
A third telephone and credit card banking centre employs 240 staff, 90% of
them female, and about 60% part-time and 40% full-time. The majority of
staff are on a contract to work flexible hours, either from 0800-2000,
or 2000-0800. The maximum contract is for 35 hours, but many staff are on
20-hour contracts, to work a shift of 4 hours on 5 days out of 7.
The company collects the top four working hour preferences from staff, and
injects that every month into a computer programme that also works out the
likely work-demand pattern, based on previous patterns and trends: about
97% of staff get either their first or second preference. This is normally done
one month ahead, and staff have freedom to swap shifts after that. It is
possible to override the programme to give staff fixed shifts, if there is some
vital personal reason for it. The computer programme is not only excellent
for time-saving in working out rotas, but is also accepted as fair by the staff. If
there are short-term changes in the apparent work-load, the company just ask
whether staff are willing to adjust to it.
A relatively small staff of three (to allow for breaks) cover the core
1600-2400 and 0000-0800 period, all working fixed shifts. The vast majority
of the work is covered on day shifts, either between 0800 and 1600 or 0900
and 1700.
Such systems do not, on the whole, result in simple printable rotas that could
be used by other companies. It is clear from these examples that many of the
choices of how to man an uneven system are in many respects based on local
considerations, and may not be easily exportable to other places.
Almost all banking systems, credit card providers, and many insurance
companies operate systems like this. Cable and satellite television providers
also offer continuously manned telephone help, partly to solve problems of
customers, and partly to stimulate sales. So do many computer hardware and
software companies. Currently telephone sales call centres are experiencing
considerable expansion in the UK.
International airlines also man 24-hour reservation systems. British Airways
operates an unusual system, using modern telephone routing systems to avoid
41
 having to man a night shift for reservations in the United Kingdom. After 10
p.m., inquiries are routed to the US.

Underlying principles
Some underlying principles can be seen in these arrangements. This is not to
say that these principles are overtly stated, but drawing them out may help to
make the lessons more generalisable:
1. The use of night shift should be minimised where this is possible.
2. If only small numbers of staff are required for night shift, it may be
possible to find people who would positively prefer to work at night.
3. If only small numbers are required at weekends, it may be possible to find
volunteers who prefer to work at weekends.
4. Management has a responsibility for working out the pattern of staff load
that will be needed. In most modern computer-controlled telephone
systems, for example, it is fairly simple to request a count of the demand
hour by hour and day by day, and estimate the regularity of recurrence of
this pattern.
5. Advance notice of working hours is highly valued by most staff.
6. It is possible to survey staff preferences and take them into account in
generating rotas.
7. A facility to exchange shifts is highly valued by most staff.
8. Consultation with staff when exceptional circumstances arise can lead to
a positive and helpful response.
Some qualifications to these principles should also be made plain:
1. Great care should be taken if staffing on night shift is reduced to a single
person. Human company is one of the best antidotes to the drowsiness
that affects most people when working at night. If the consequences of
error are anything more than trivial, at least two people should always be
there, to allow for personal natural breaks of one kind and another.
2. The use of permanent fixed night shift staff can easily lead to less than
adequate communication with daytime management, and steps may need
to be taken to counteract this.
3. It is important that a set of principles be developed and adhered to that is
both efficient at meeting demand and fair and humane in dealing with
human operators, with important aspects of their lives outside work.

42
Chapter 6: CONCLUSIONS

The prevalence figures for shiftwork in general support the impression that
shiftwork is on a slow but steady increase, probably as a result of increasing
world-wide competitiveness.
Even more interesting is the impression that continuous shiftwork is
increasingly occurring in sectors where it was formerly uncommon: financial
services, most obviously, but also many working situations where there is
pressure for longer operating time, to spread out the overheads of capital
investment, or to provide continuous customer service.
Often these business pressures require considerable flexibility in response, and
achieving this without damaging the lives of the staff involved is a major
challenge for management. How can an uneven work demand be shaped to
provide some kind of regularity and predictability in working time schedules?
A further stimulus to change in countries of the European Union has been
the steady impact of the EU Directive on the Organisation of Working Time.
This required member states to review their national legislation on working
time by November 1996, three years after the Directive was approved. This
target for completion has been met in some states, and has stimulated
consultation and discussion in all states. Not all states had previous
legislation that could provide a framework for change. In the case of the
United Kingdom, where there was no existing legislation, the government
raised a case against the validity of the Directive, and only began the
consultative process after this case was lost in 1996; a general election in
1997, and the change to a new government with many other priorities for
inclusion in the legislative programme, have caused further delays. In several
other states – Ireland, for example – the legislation has been prepared, but
has not yet come into force. In general, the directive is likely to continue to
have a major impact on working time in the European Union, and beyond as
other countries watch these developments with interest.
All these stimuli to creative and intelligent organisations, trying to adapt
lessons from the past to their own particular situations, locations and work
forces, generate continuing questions and interest in approaches to solving
shiftwork problems. We hope that this collection helps to contribute to this
process.

43
References

Bauer, F., Groß, H. And Schilling, G. (1994) Arbeitszeit ’93. Arbeitszeiten,

Arbeitszeitwünsche, Zeitbewirtschaftung, Arbeitszeitgestaltungschancen von
abhängig Beschäftigten (Working hours, working time preferences, time
management, and dependent employees’ opportunities to organise their time
as desired). Institut zur Erforschung sozialer Chancen, Köln (1994).
Ministerium für Arbeit, Gesundheit und Soziales des Landes Nordrhein-
Westfalen (ed.)
BEST 7 (1994) Instruments for the design, implementation and assessment of
working time arrangements. (Dublin: European Foundation for the
Improvement of Living and Working Conditions)
Blanpain, R., Köhler, E. and Rojot, J. (eds) (1997) Legal and contractual
limitations to working time in the European Union. (Dublin: European
Foundation for the Improvement of Living and Working Conditions)
Bloch-London, C. (1992). Le travail en équipes successives (Shiftwork).
Dossiers Statistiques du Travail et de l’Emploi, 1992, 85, pp. 41-82.
CBS (1996) Centraal Bureau voor de Statistieken (Central Bureau for Statistics).
Enquete beroepsbevolking (Labour Force Survey) 1995, Heerlen.
Costa, G. (1995) Some examples of arrangements of continuous shift systems
in Italy. (Dublin: European Foundation for the Improvement of Living and
Working Conditions)
Dorel, M. (1995) Roulement en horaires alternés de travail pour les équipes
de conduite de centrales nucléaires (Rotation in alternating hours of work for
control teams at nuclear centres). Rapport interne, Laboratoire Travail et
Cognition, Université de Toulouse le Mirail.
Dorel, M., de Terssac, G., Thon, P., and Quéinnec, Y. (1984) Nouveauté
technique, restructuration et travail en équipes successives dan la sidérurgie
française (New technology, restructuring and shiftwork in the French steel
industry). ADEACT, Toulouse.
Dorel, M. (1994) Temps alternés de travail, Activité en poste mobile
(Changing times of work: Activity in a mobile station). Laboratoire Travail
et Cognition, Université de Toulouse Le Mirail.
FNV (1992) Vernieuwing van roosters (Renewal of rosters). (Amsterdam:
Industriebond FNV)
Gadbois, C. (1981) Aides-soignantes et infirmières de nuit (Care assistants
and night nurses). Conditions de Travail et Vie Quotidienne, ANACT, 1981,
77p.
Gadbois, C. and Dorel, M. (1996) Continuous shiftwork: Various practices
for different technical and social needs. (Dublin: European Foundation for
the Improvement of Living and Working Conditions)
Gadbois, C., and Prunier, S. (1994) Douaniers en horaires atypiques: rythmes
de vie et conditions de travail (Customs officers on atypical working hours:
rhythms of life and conditions of work). Rapport de recherche, 95 p. +
annexes (EPHE, Paris)
44
Gadbois, C., and Prunier, S. (1995) Shiftwork and job constraints in French
customs: a survey of a round-the-clock occupation. Shiftwork International
Newsletter, May 1995, 12, 1, p. 110.
Guggenbuhl, N. (1993) Alimentation et cancer (Food and cancer). Brussels
Association contre le cancer.
Haumont, P. and Gobert, R. (1979) Description of 6-shift system in power
station. (Dublin: European Foundation for the Improvement of Living and
Working Conditions)
Hornberger, S. (1994) Auswirkungen arbeitswissenschaftlich verbesserter
Schichtsysteme auf die Schichtarbeiter. Eine kontrollierte
Längsschnittuntersuchung mit 5 Experimentalgruppen (The effects of
scientifically improved shift systems on shiftworkers. A controlled
longitudinal study of 5 experimental groups). (Peter Lang, Frankfurt am
Main, Bern, New York, Paris)
Hornberger, S. and Knauth, P. (1995) Effects of various types of change in
shift schedules: a controlled longitudinal study. Work and Stress: Night and
Shiftwork, 9, No. 2/3, pp.124-133.
Knauth, P. (1993) The design of shift systems. Ergonomics, 36, 1-3, pp.15-28.
Knauth, P. (1996) Design of shiftwork systems. In W.P. Colquhoun, G. Costa,
S. Folkard and P. Knauth, Shiftwork: Problems and Solutions, chapter 7. (Peter
Lang, Frankfurt am Main, Bern, New York, Paris)
Knauth, P. and Hornberger, S. (1992) Instruments for the design,
implementation and assessment of working time arrangements. (Dublin:
European Foundation for the Improvement of Living and Working
Conditions)
Knauth, P. and Hornberger, S. (1995) Continuous shift systems. (Dublin:
European Foundation for the Improvement of Living and Working
Conditions)
K.U.L. (1980) The desirable and possible management-options regarding
shiftwork. (Higher Institute for Labour).
National Institute of Statistics, Belgium (1994) The importance of shiftwork
in Belgium.
Léonard, R. (1995) Continuous shift systems. (Dublin: European Foundation
for the Improvement of Living and Working Conditions)
Léonard, R. (1981) Night and shiftwork. In A. Reinberg, N. Vieux and P.
Andlauer (eds.), Night and Shiftwork: Biological and Social Aspects, pp.323-329.
(Pergamon Press, Oxford)
Léonard, R. and Brogniaux, L. (1986) Survey on shiftwork in Belgian
hospitals: review of work conditions. ONEM (National Service for
Employment and Labour), 23 May 1986.
Léonard, R. and Vandeville, V. (1994) Analysis of shiftwork in the industry
and services. National Service for Employment and Labour, Brussels.
Mairiaux, P. (1995) Shiftwork and biological rhythms. Lecture course notes,
Université catholique de Louvain, Department of Occupational Medicine,
1986-1995.
Prunier, S., and Gadbois, C. (1995) “Six-hour shiftwork: unfavourable
consequences of a split workweek.” Shiftwork International Newsletter, May
1995, 12, 1, p. 110.
45
 Ravasio, B. and Zanzottera, E. (1991) Gli orari e l’utilizzo degli impianti
(Working hours and plants utilisation). Contrattare il tempo (Negotiating
the working hours), Datanews Ed. Srl., Rome, pp.124-133.
Rosa, R. and Bonnet, M.H. (1993) Performance and alertness on 8 h and
12 h rotating shifts at a natural gas utility. Ergonomics, 36, 10, pp. 1177-1193.
Van der Weerd, E.C. and Jansen, B. (1996) Recent trends in round-the-clock
systems in the Netherlands. (Dublin: European Foundation for the
Improvement of Living and Working Conditions)

46
47 47
 EF/96/••/EN
48 48
 What is BEST?

BEST BEST is the Bulletin of European Studies on Time,

offering condensed and practical information on important

4
developments in the field of work organisation and shiftwork.
BEST is published twice yearly in English, French and

5
German.

SY-AB-98-001-EN-C
Who is BEST?
A European network of experts, set up by the European
Foundation for the Improvement of Living and
Working Conditions to monitor developments, is responsible
for the content of the bulletin.
Members of the network:
• Giovanni COSTA
• Charles GADBOIS
• Ben JANSEN
• Peter KNAUTH
• Robert LÉONARD
• Alexander WEDDERBURN

Whom does BEST

address?
The bulletin is aimed at those who are interested in
keeping abreast of developments in the field of work organisation
and shiftwork in the Member States of the Community. It
especially addresses: decision-makers at
policy level; organisations representing employers and
workers; managers and workers within firms.

Suggestions? Comments?
Enquiries? Further copies?
Write to:
Dimitrios POLITIS or Pascal PAOLI at the
European Foundation for the Improvement of Living and Working
Conditions, Wyattville Road, Loughlinstown,
Co. Dublin, Ireland. Tel: +353 1 204 3100.
or to:
Alexander WEDDERBURN, Department of Business
Organisation, School of Management, Heriot-Watt University,
Edinburgh EH14 4AT, United Kingdom,
Tel: +44 131 449 5111. Fax: +44 131 451 3296
Email: [email protected]

The Bulletin of European Studies on Time is published under the auspices of the European Foundation
for the Improvement of Living and Working Conditions. The opinions expressed in it are those of the
editorial board and do not necessarily reflect the opinion of the Foundation.

EF/98/02/EN