11/16/2020

Chapter 14

Waiting Lines and Queuing

Theory Models

To accompany
Quantitative Analysis for Management, Tenth Edition,
by Render, Stair, and Hanna © 2008 Prentice-Hall, Inc.
Power Point slides created by Jeff Heyl © 2009 Prentice-Hall, Inc.

Learning Objectives
After completing this chapter, students will be able to:
1. Describe the trade-off curves for cost-of-
waiting time and cost-of-service
2. Understand the three parts of a queuing
system: the calling population, the queue
itself, and the service facility
3. Describe the basic queuing system
configurations
4. Understand the assumptions of the common
models dealt with in this chapter
5. Analyze a variety of operating characteristics
of waiting lines
© 2009 Prentice-Hall, Inc. 14 – 2

1
 11/16/2020

Chapter Outline
14.1 Introduction
14.2 Waiting Line Costs
14.3 Characteristics of a Queuing System
14.4 Single-Channel Queuing Model with Poisson Arrivals and
Exponential Service Times (M/M/1)
14.5 Multichannel Queuing Model with Poisson Arrivals and
Exponential Service Times (M/M/m)
14.6 Constant Service Time Model (M/D/1)
14.7 Finite Population Model (M/M/1 with Finite Source)
14.8 Some General Operating Characteristic Relationships
14.9 More Complex Queuing Models and the Use of
Simulation

© 2009 Prentice-Hall, Inc. 14 – 3

Introduction
◼ Queuing theory is the study of waiting lines
◼ It is one of the oldest and most widely used
quantitative analysis techniques
◼ Waiting lines are an everyday occurrence for
most people
◼ Queues form in business process as well
◼ The three basic components of a queuing
process are arrivals, service facilities, and the
actual waiting line
◼ Analytical models of waiting lines can help
managers evaluate the cost and effectiveness
of service systems

© 2009 Prentice-Hall, Inc. 14 – 4

2
 11/16/2020

Waiting Line Costs

◼ Most waiting line problems are focused on
finding the ideal level of service a firm should
provide
◼ In most cases, this service level is something
management can control
◼ When an organization does have control, they
often try to find the balance between two
extremes
◼ A large staff and many service facilities generally
results in high levels of service but have high
costs

© 2009 Prentice-Hall, Inc. 14 – 5

Waiting Line Costs

◼ Having the minimum number of service facilities
keeps service cost down but may result in
dissatisfied customers
◼ There is generally a trade-off between cost of
providing service and cost of waiting time
◼ Service facilities are evaluated on their total
expected cost which is the sum of service costs
and waiting costs
◼ Organizations typically want to find the service
level that minimizes the total expected cost

© 2009 Prentice-Hall, Inc. 14 – 6

3
 11/16/2020

Waiting Line Costs

◼ Queuing costs and service level

Total Expected Cost

Cost of Providing Service

Cost

Cost of Waiting Time

* Service Level
Optimal
Figure 14.1 Service
Level
© 2009 Prentice-Hall, Inc. 14 – 7

Three Rivers Shipping Company

Example
◼ Three Rivers Shipping operates a docking facility
on the Ohio River
◼ An average of 5 ships arrive to unload their
cargos each shift
◼ Idle ships are expensive
◼ More staff can be hired to unload the ships, but
that is expensive as well
◼ Three Rivers Shipping Company wants to
determine the optimal number of teams of
stevedores to employ each shift to obtain the
minimum total expected cost

© 2009 Prentice-Hall, Inc. 14 – 8

4
 11/16/2020

Three Rivers Shipping Company

Example
◼ Three Rivers Shipping waiting line cost analysis
NUMBER OF TEAMS OF STEVEDORES
WORKING
1 2 3 4
(a) Average number of ships arriving 5 5 5 5
per shift
(b) Average time each ship waits to 7 4 3 2
be unloaded (hours)
(c) Total ship hours lost per shift 35 20 15 10
(a x b)
(d) Estimated cost per hour of idle $1,000 $1,000 $1,000 $1,000
ship time
(e) Value of ship’s lost time or waiting $35,000 $20,000 $15,000 $10,000
cost (c x d)
(f) Stevedore team salary or service $6,000 $12,000 $18,000 $24,000
cost
(g) Total expected cost (e + f) $41,000 $32,000 $33,000 $34,000
Optimal cost
Table 14.1
© 2009 Prentice-Hall, Inc. 14 – 9

Characteristics of a Queuing System

◼ There are three parts to a queuing system

1. The arrivals or inputs to the system
(sometimes referred to as the calling
population)
2. The queue or waiting line itself
3. The service facility
◼ These components have their own
characteristics that must be examined
before mathematical models can be
developed

© 2009 Prentice-Hall, Inc. 14 – 10

10

5
 11/16/2020

Characteristics of a Queuing System

◼ Arrival Characteristics have three major

characteristics, size, pattern, and behavior
◼ Size of the calling population
◼ Can be either unlimited (essentially infinite)
or limited (finite)
◼ Pattern of arrivals
◼ Can arrive according to a known pattern or
can arrive randomly
◼ Random arrivals generally follow a Poisson
distribution

© 2009 Prentice-Hall, Inc. 14 – 11

11

Characteristics of a Queuing System

◼ The Poisson distribution is

e −  X
P( X ) = for X = 0, 1, 2, 3, 4,...
X!
where
P(X) = probability of X arrivals
X = number of arrivals per unit of time
 = average arrival rate
e = 2.7183

© 2009 Prentice-Hall, Inc. 14 – 12

12

6
 11/16/2020

Characteristics of a Queuing System

◼ If  = 2 parts/min, we can find the values for

X = 0, 1, and 2

e −  X
P( X ) =
X!
e − 2 20 0.1353(1)
P (0 ) = = = 0.1353 = 14%
0! 1
e − 2 21 e − 2 2 0.1353(2)
P (1) = = = = 0.2706 = 27%
1! 1 1
e − 2 22 e − 2 4 0.1353( 4 )
P (2 ) = = = = 0.2706 = 27%
2! 2(1) 2
© 2009 Prentice-Hall, Inc. 14 – 13

13

Characteristics of a Queuing System

◼ Two examples of the Poisson distribution for

arrival rates

0.30 – 0.25 –
0.25 – 0.20 –
0.20 –
Probability

Probability

0.15 –
0.15 –
0.10 –
0.10 –
0.05 – 0.05 –

0.00 –| | | | | | | | | | 0.00 –| | | | | | | | | |
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9
X X
 = 2 Distribution  = 4 Distribution

© 2009 Prentice-Hall, Inc. 14 – 14

14

7
 11/16/2020

Characteristics of a Queuing System

◼ Behavior of arrivals
◼ Most queuing models assume customers are
patient and will wait in the queue until they are
served and do not switch lines
◼ Balking refers to customers who refuse to join
the queue
◼ Reneging customers enter the queue but
become impatient and leave without receiving
their service
◼ That these behaviors exist is a strong
argument for the use of queuing theory to
managing waiting lines

© 2009 Prentice-Hall, Inc. 14 – 15

15

Characteristics of a Queuing System

◼ Waiting Line Characteristics

◼ Waiting lines can be either limited or unlimited
capacity
◼ Queue discipline refers to the rule by which
customers in the line receive service
◼ The most common rule is first-in, first-out
(FIFO)
◼ Other rules are possible and may be based on
other important characteristics
◼ Other rules can be applied to select which
customers enter which queue, but may apply
FIFO once they are in the queue
© 2009 Prentice-Hall, Inc. 14 – 16

16

8
 11/16/2020

Characteristics of a Queuing System

◼ Service Facility Characteristics

◼ Basic queuing system configurations
◼ Service systems are classified in terms of
the number of channels, or servers, and the
number of phases, or service stops
◼ A single-channel system with one server is
quite common
◼ Multichannel systems exist when multiple
servers are fed by one common waiting line
◼ In a single-phase system the customer
receives service form just one server
◼ If a customer has to go through more than
one server, it is a multiphase system
© 2009 Prentice-Hall, Inc. 14 – 17

17

Characteristics of a Queuing System

◼ Four basic queuing system configurations

Queue

Service Departures
Arrivals Facility after Service

Single-Channel, Single-Phase System

Queue
Type 1 Type 2 Departures
Arrivals Service Service after
Facility Facility Service

Single-Channel, Multiphase System

© 2009 Prentice-Hall, Inc. 14 – 18

18

9
 11/16/2020

Characteristics of a Queuing System

◼ Four basic queuing system configurations

Service
Facility Departures
Queue
1

Service
Arrivals Facility after
2

Service
Facility Service
3

Multichannel, Single-Phase System

© 2009 Prentice-Hall, Inc. 14 – 19

19

Characteristics of a Queuing System

◼ Four basic queuing system configurations

Type 1 Type 2
Queue Service Service
Facility Facility
1 1 Departures
Arrivals after
Service
Type 1 Type 2
Service Service
Facility Facility
2 2

Multichannel, Multiphase System

© 2009 Prentice-Hall, Inc. 14 – 20

20

10
 11/16/2020

Characteristics of a Queuing System

◼ Service time distribution

◼ Service patterns can be either constant or
random
◼ Constant service times are often machine
controlled
◼ More often, service times are randomly
distributed according to a negative
exponential probability distribution
◼ Models are based on the assumption of
particular probability distributions
◼ Analysts should take to ensure observations
fit the assumed distributions when applying
these models
© 2009 Prentice-Hall, Inc. 14 – 21

21

Characteristics of a Queuing System

◼ Two examples of exponential distribution for

service times
f (x)
– f (x) = e– x
for x ≥ 0
– and  > 0
–

–  = Average Number Served per Minute

–
Average Service Time
– of 20 Minutes

–
Average Service Time of 1 Hour
|– | | | | | |
0 30 60 90 120 150 180 Service Time (Minutes)
© 2009 Prentice-Hall, Inc. 14 – 22

22

11
 11/16/2020

Identifying Models Using

Kendall Notation
◼ D. G. Kendall developed a notation for queuing
models that specifies the pattern of arrival, the
service time distribution, and the number of
channels
◼ It is of the form
Arrival Service time Number of service
distribution distribution channels open

◼ Specific letters are used to represent probability

distributions
M = Poisson distribution for number of occurrences
D = constant (deterministic) rate
G = general distribution with known mean and variance

© 2009 Prentice-Hall, Inc. 14 – 23

23

Identifying Models Using

Kendall Notation
◼ So a single channel model with Poisson arrivals
and exponential service times would be
represented by
M /M / 1
◼ If a second channel is added we would have
M /M / 2
◼ A three channel system with Poisson arrivals and
constant service time would be
M /D / 3
◼ A four channel system with Poisson arrivals and
normally distributed service times would be
M /G / 4

© 2009 Prentice-Hall, Inc. 14 – 24

24

12
 11/16/2020

Single-Channel Model, Poisson Arrivals,

Exponential Service Times (M/M/1)
◼ Assumptions of the model
◼ Arrivals are served on a FIFO basis
◼ No balking or reneging
◼ Arrivals are independent of each other but rate is constant
over time
◼ Arrivals follow a Poisson distribution
◼ Service times are variable and independent but the average
is known
◼ Service times follow a negative exponential distribution
◼ Average service rate is greater than the average arrival rate

© 2009 Prentice-Hall, Inc. 14 – 25

25

Single-Channel Model, Poisson Arrivals,

Exponential Service Times (M/M/1)
◼ When these assumptions are met, we can
develop a series of equations that define the
queue’s operating characteristics

◼ Queuing Equations
◼ We let
 = mean number of arrivals per time period
 = mean number of people or items served
per time period
◼ The arrival rate and the service rate must be
for the same time period

© 2009 Prentice-Hall, Inc. 14 – 26

26

13
 11/16/2020

Single-Channel Model, Poisson Arrivals,

Exponential Service Times (M/M/1)
1. The average number of customers or units in the
system, L 
L=
−
2. The average time a customer spends in the
system, W 1
W =
−
3. The average number of customers in the queue, Lq
2
Lq =
( −  )

© 2009 Prentice-Hall, Inc. 14 – 27

27

Single-Channel Model, Poisson Arrivals,

Exponential Service Times (M/M/1)
4. The average time a customer spends waiting in
the queue, Wq

Wq =
( −  )

5. The utilization factor for the system, , the

probability the service facility is being used

=


© 2009 Prentice-Hall, Inc. 14 – 28

28

14
 11/16/2020

Single-Channel Model, Poisson Arrivals,

Exponential Service Times (M/M/1)
6. The percent idle time, P0, the probability no one is
in the system

P0 = 1 −


7. The probability that the number of customers in

the system is greater than k, Pn>k
k +1

Pn k =  


© 2009 Prentice-Hall, Inc. 14 – 29

29

Arnold’s Muffler Shop Case

◼ Arnold’s mechanic can install mufflers at a rate of

3 per hour
◼ Customers arrive at a rate of 2 per hour
 = 2 cars arriving per hour
 = 3 cars serviced per hour

 2 2
L= = = = 2 cars in the system
− 3−2 1 on the average

1 1
W = = = 1 hour that an average car
 − 3−2 spends in the system

© 2009 Prentice-Hall, Inc. 14 – 30

30

15
 11/16/2020

Arnold’s Muffler Shop Case

2 22 4
Lq = = = = 1.33 cars waiting in line
 (  −  ) 3(3 − 2) 3(1) on the average
 2
Wq = = hour = 40 minutes average
( −  ) 3 waiting time per car
 2
= = = 0.67 = percentage of time
 3 mechanic is busy
 2
P0 = 1 − = 1 − = 0.33 = probability that there
 3 are 0 cars in the system

© 2009 Prentice-Hall, Inc. 14 – 31

31

Arnold’s Muffler Shop Case

◼ Probability of more than k cars in the system

k Pn>k = (2/3)k+1

0 0.667 Note that this is equal to 1 – P0 = 1 – 0.33 = 0.667

1 0.444
2 0.296
Implies that there is a 19.8% chance that
3 0.198
more than 3 cars are in the system
4 0.132
5 0.088
6 0.058
7 0.039

© 2009 Prentice-Hall, Inc. 14 – 32

32

16
 11/16/2020

Arnold’s Muffler Shop Case

◼ Arnold is thinking about hiring a different

mechanic who can install mufflers at a faster rate
◼ The new operating characteristics would be
 = 2 cars arriving per hour
 = 4 cars serviced per hour

 2 2
L= = = = 1 car in the system
− 4−2 2 on the average

1 1
W = = = 1/2 hour that an average car
 − 4−2 spends in the system

© 2009 Prentice-Hall, Inc. 14 – 33

33

Arnold’s Muffler Shop Case

2 22 4
Lq = = = = 1/2 cars waiting in line
 (  −  ) 4( 4 − 2) 8(1) on the average
 1
Wq = = hour = 15 minutes average
( −  ) 4 waiting time per car
 2
= = = 0.5 = percentage of time
 4 mechanic is busy
 2
P0 = 1 − = 1 − = 0.5 = probability that there
 4 are 0 cars in the system

© 2009 Prentice-Hall, Inc. 14 – 34

34

17
 11/16/2020

Arnold’s Muffler Shop Case

◼ Probability of more than k cars in the system

k Pn>k = (2/4)k+1

0 0.500
1 0.250
2 0.125
3 0.062

4 0.031
5 0.016
6 0.008
7 0.004

© 2009 Prentice-Hall, Inc. 14 – 35

35

Multichannel Model, Poisson Arrivals,

Exponential Service Times (M/M/m)
◼ Assumptions of the model
◼ Arrivals are served on a FIFO basis
◼ No balking or reneging
◼ Arrivals are independent of each other but rate
is constant over time
◼ Arrivals follow a Poisson distribution
◼ Service times are variable and independent but
the average is known
◼ Service times follow a negative exponential
distribution
◼ Average service rate is greater than the
average arrival rate
© 2009 Prentice-Hall, Inc. 14 – 36

36

18
 11/16/2020

Multichannel Model, Poisson Arrivals,

Exponential Service Times (M/M/m)
◼ Equations for the multichannel queuing model
◼ We let
m = number of channels open
 = average arrival rate
 = average service rate at each channel
1. The probability that there are zero customers in
the system
1
P0 = for m   
 n = m −1
1  1 
n m
m
    +  
n!     m!    m  − 
 n=0

© 2009 Prentice-Hall, Inc. 14 – 37

37

Multichannel Model, Poisson Arrivals,

Exponential Service Times (M/M/m)
2. The average number of customer in the system

 ( /  )m 
L= P +
( m − 1)!( m −  ) 2 0


3. The average time a unit spends in the waiting line

or being served, in the system

 ( /  )m 1 L
W= P + =
( m − 1)!( m −  ) 2 0
 

© 2009 Prentice-Hall, Inc. 14 – 38

38

19
 11/16/2020

Multichannel Model, Poisson Arrivals,

Exponential Service Times (M/M/m)
4. The average number of customers or units in line
waiting for service

Lq = L −

5. The average number of customers or units in line
waiting for service
1 Lq
Wq = W − =
 
6. The average number of customers or units in line
waiting for service

=
m
© 2009 Prentice-Hall, Inc. 14 – 39

39

Arnold’s Muffler Shop Revisited

◼ Arnold wants to investigate opening a second
garage bay
◼ He would hire a second worker who works at the
same rate as his first worker
◼ The customer arrival rate remains the same

1
P0 = for m   
 n = m −1 1    n  1    m m 
    +  
 n = 0 n!     m!    m  − 

P0 = 0.5
= probability of 0 cars in the system

© 2009 Prentice-Hall, Inc. 14 – 40

40

20
 11/16/2020

Arnold’s Muffler Shop Revisited

◼ Average number of cars in the system

 ( /  )m 
L= P + = 0.75
( m − 1)!( m −  ) 2 0


◼ Average time a car spends in the system

L 3
W = = hours = 22 1 minutes
 8 2

© 2009 Prentice-Hall, Inc. 14 – 41

41

Arnold’s Muffler Shop Revisited

◼ Average number of cars in the queue

 3 2 1
Lq = L − = − = = 0.083
 4 3 12

◼ Average time a car spends in the queue

1 Lq 0.083
Wq = W − = = = 0.0415 hour = 2 1 minutes
  2 2

© 2009 Prentice-Hall, Inc. 14 – 42

42

21
 11/16/2020

Arnold’s Muffler Shop Revisited

◼ Effect of service level on Arnold’s operating
characteristics
LEVEL OF SERVICE
ONE TWO ONE FAST
OPERATING MECHANIC MECHANICS MECHANIC
CHARACTERISTIC =3  = 3 FOR BOTH =4
Probability that the system
0.33 0.50 0.50
is empty (P0)
Average number of cars in
2 cars 0.75 cars 1 car
the system (L)
Average time spent in the
60 minutes 22.5 minutes 30 minutes
system (W)
Average number of cars in
1.33 cars 0.083 car 0.50 car
the queue (Lq)
Average time spent in the
40 minutes 2.5 minutes 15 minutes
queue (Wq)
Table 14.2
© 2009 Prentice-Hall, Inc. 14 – 43

43

Constant Service Time Model (M/D/1)

◼ Constant service times are used when

customers or units are processed
according to a fixed cycle
◼ The values for Lq, Wq, L, and W are always
less than they would be for models with
variable service time
◼ In fact both average queue length and
average waiting time are halved in
constant service rate models

© 2009 Prentice-Hall, Inc. 14 – 44

44

22
 11/16/2020

Constant Service Time Model (M/D/1)

1. Average length of the queue

2
Lq =
2 (  −  )

2. Average waiting time in the queue


Wq =
2 (  −  )

© 2009 Prentice-Hall, Inc. 14 – 45

45

Constant Service Time Model (M/D/1)

3. Average number of customers in the system


L = Lq +


4. Average time in the system

1
W = Wq +


© 2009 Prentice-Hall, Inc. 14 – 46

46

23
 11/16/2020

Constant Service Time Model (M/D/1)

◼ Garcia-Golding Recycling, Inc.

◼ The company collects and compacts aluminum
cans and glass bottles
◼ Trucks arrive at an average rate of 8 per hour
(Poisson distribution)
◼ Truck drivers wait about 15 min before they
empty their load
◼ Drivers and trucks cost $60 per hour
◼ New automated machine can process
truckloads at a constant rate of 12 per hour
◼ New compactor will be amortized at $3 per
truck

© 2009 Prentice-Hall, Inc. 14 – 47

47

Constant Service Time Model (M/D/1)

◼ Analysis of cost versus benefit of the purchase
Current waiting cost/trip = (1/4 hour waiting time)($60/hour cost)
= $15/trip
New system:  = 8 trucks/hour arriving
 = 12 trucks/hour served
Average waiting
time in queue = Wq = 1/12 hour
Waiting cost/trip
with new compactor = (1/12 hour wait)($60/hour cost) = $5/trip
Savings with
new equipment = $15 (current system) – $5 (new system)
= $10 per trip
Cost of new equipment
amortized = $3/trip
Net savings = $7/trip
© 2009 Prentice-Hall, Inc. 14 – 48

48

24
 11/16/2020

Finite Population Model

(M/M/1/N with Finite Source)
◼ When the population of potential customers is
limited, the models are different
◼ There is now a dependent relationship between
the length of the queue and the arrival rate
◼ The model has the following assumptions
1. There is only one server
2. The population of units seeking service is
finite
3. Arrivals follow a Poisson distribution and
service times are exponentially distributed
4. Customers are served on a first-come, first-
served basis

© 2009 Prentice-Hall, Inc. 14 – 49

49

Finite Population Model

(M/M/1/N with Finite Source)

◼ Equations for the finite population model

◼ Using
 = mean arrival rate,  = mean service rate,
N = size of the population
◼ The operating characteristics are

1. Probability that the system is empty

1
P0 =
N!   
N n

  
n = 0 ( N − n )!   

© 2009 Prentice-Hall, Inc. 14 – 50

50

25
 11/16/2020

Finite Population Model

(M/M/1/N with Finite Source)
2. Average length of the queue
 + 
Lq = N −  (1 − P0 )
  

3. Average number of customers (units) in the system

L = Lq + (1− P0 )

4. Average waiting time in the queue

Lq
Wq =
( N − L)

© 2009 Prentice-Hall, Inc. 14 – 51

51

Finite Population Model

(M/M/1/N with Finite Source)
5. Average time in the system

1
W = Wq +


6. Probability of n units in the system

N!   
n

Pn =   P for n = 0,1,..., N
( N − n )!    0

© 2009 Prentice-Hall, Inc. 14 – 52

52

26
 11/16/2020

Department of Commerce Example

◼ The Department of Commerce has five printers

that each need repair after about 20 hours of work
◼ Breakdowns follow a Poisson distribution
◼ The technician can service a printer in an average
of about 2 hours, following an exponential
distribution
 = 1/20 = 0.05 printer/hour
 = 1/2 = 0.50 printer/hour

© 2009 Prentice-Hall, Inc. 14 – 53

53

Department of Commerce Example

1. 1
P0 = n
= 0.564
5
5!  0.05 
  
n = 0 (5 − n )!  0.5 

2.  0.05 + 0.5 
Lq = 5 −  (1 − P0 ) = 0.2 printer
 0.05 

3. L = 0.2 + (1− 0.564 ) = 0.64 printer

© 2009 Prentice-Hall, Inc. 14 – 54

54

27
 11/16/2020

Department of Commerce Example

4. 0.2 0.2
Wq = = = 0.91 hour
(5 − 0.64 )(0.05 ) 0.22

1
5. W = 0.91 + = 2.91 hours
0.50
◼ If printer downtime costs $120 per hour and the
technician is paid $25 per hour, the total cost is
Total (Average number of printers down)
hourly = (Cost per downtime hour)
cost + Cost per technician hour

= (0.64)($120) + $25 = $101.80

© 2009 Prentice-Hall, Inc. 14 – 55

55

Some General Operating

Characteristic Relationships
◼ Certain relationships exist among specific
operating characteristics for any queuing system
in a steady state
◼ A steady state condition exists when a system is
in its normal stabilized condition, usually after an
initial transient state
◼ The first of these are referred to as Little’s Flow
Equations
L = W (or W = L/ )
Lq =  Wq (or Wq = Lq/ )
◼ And

W = Wq + 1/
© 2009 Prentice-Hall, Inc. 14 – 56

56

28
 11/16/2020

More Complex Queuing Models and

the Use of Simulation
◼ In the real world there are often variations from
basic queuing models
◼ Computer simulation can be used to solve these
more complex problems
◼ Simulation allows the analysis of controllable
factors
◼ Simulation should be used when standard
queuing models provide only a poor
approximation of the actual service system

© 2009 Prentice-Hall, Inc. 14 – 57

57

29