Ch a pt e r 2 5 Pa ge 1

5/ 30/ 02
Inventory Theory
Inventories are materials stored, waiting for processing, or experiencing processing.
They are ubiquitous throughout all sectors of the economy. Observation of almost any
company balance sheet, for example, reveals that a significant portion of its assets
comprises inventories of raw materials, components and subassemblies within the
production process, and finished goods. Most managers don't like inventories because
they are like money placed in a drawer, assets tied up in investments that are not
producing any return and, in fact, incurring a borrowing cost. They also incur costs for
the care of the stored material and are subject to spoilage and obsolescence. In the last
two decades their have been a spate of programs developed by industry, all aimed at
reducing inventory levels and increasing efficiency on the shop floor. Some of the most
popular are conwip, kanban, just-in-time manufacturing, lean manufacturing, and flexible
manufacturing.
Nevertheless, in spite of the bad features associated with inventories, they do have
positive purposes. Raw material inventories provide a stable source of input required for
production. A large inventory requires fewer replenishments and may reduce ordering
costs because of economies of scale. In-process inventories reduce the impacts of the
variability of the production rates in a plant and protect against failures in the processes.
Final goods inventories provide for better customer service. The variety and easy
availability of the product is an important marketing consideration. There are other kinds
of inventories, including spare parts inventories for maintenance and excess capacity built
into facilities to take advantage of the economies of scale of construction.
Because of their practical and economic importance, the subject of inventory
control is a major consideration in many situations. Questions must be constantly
answered as to when and how much raw material should be ordered, when a production
order should be released to the plant, what level of safety stock should be maintained at a
retail outlet, or how in-process inventory is to be maintained in a production process.
These questions are amenable to quantitative analysis with the help of inventory theory.
25.1 Inventory Models
In this chapter, we will consider several types of models starting with the deterministic
case in the next section. Even though many features of an inventory system involve
uncertainty of some kind, it is common to assume much simpler deterministic models for
which solutions are found using calculus. Deterministic models also provide a base on
which to incorporate assumptions concerning uncertainty. Section 25.3 adds a stochastic
dimension to the model with random product demand. Section 25.4 begins discussion of
stochastic inventory systems with the single period stochastic model. The model has
applications for products for which the ordering process is nonrepeating. The remainder
of the chapter addresses models with an infinite time horizon and several assumptions
2 Inventory Theory
regarding the costs of operation. Sections 25.5 and 25.6 derive optimal solutions for the
(s, S) policy under a variety of conditions. This policy places an order up to level S when
the inventory level falls to the reorder point s. Section 25.7 extends these results to the
(R, S) policy. In this case, the inventory is observed periodically (with a time interval R),
and is replenished to level S.
Flow, Inventory and Time
An inventory is represented in the simple diagram of Fig. 1. Items flow
into the system, remain for a time and then flow out. Inventories occur
whenever the time an individual enters is different than when it leaves.
During the intervening interval the item is part of the inventory.
Flow In
Inventory Level
(Residence Time)
Flow Out
Figure 1. A system component with inventory
For example, say the box in Fig. 1 represents a manufacturing
process that takes a fixed amount of time. A product entering the box at
one moment leaves the box one hour later. Products arrive at a rate of 100
per hour. Clearly, if we look in the box, we will find some number of
items. That number is the inventory level. The relation between flow,
time and inventory level that is basic to all systems is
Inventory level = (Flow rate )(Residence time) (1)
where the flow rate is expressed in the same time units as the residence
time. For the example, we have
Inventory Level = (100 products/hour )(1 hour) = 100 products.
When the factors in Eq. (1) are not constant in time, we typically use their
mean values.
Whenever two of the factors in the above expression are given, the
third is easily computed. Consider a queueing system for which customers
are observed to arrive at an average rate of 10 per hour. When the
customer finds the servers busy, he or she must wait. Customers in the
system, either waiting or be served, are the inventory for this system.
Using a sampling procedure we determine that the average number of
customers in the inventory is 5. We ask, how long on the average is each
customer in the system? Using the relation between the flow, time and
Inventory Models 3
inventory, we determine the answer as 0.5 hours. As we saw in the
Chapter 16, Queueing Models, Eq (1) is called Little's Law.
The relation between time and inventory is significant, because
very often reducing the throughput time for a system is just as important
as reducing the inventory level. Since they are proportional, changing one
factor inevitably changes the other.
The Inventory Level
The inventory level depends on the relative rates of flow in and out of the
system. Define y(t) as the rate of input flow at time t and Y(t) the
cumulative flow into the system. Define z(t) as the rate of output flow at
time t and Z(t) as the cumulative flow out of the system. The inventory
level, I(t) is the cumulative input less the cumulative output.
I(t) = Y(t) – Z(t) =


0
t
y(x)dx -


0
t
z(x)dx (2)
Figure 2 represents the inventory for a system when the rates vary with
time.
Time
Inventory Level
0
0
Figure 2. Inventory fluctuations as a function of time
The figure might represent a raw material inventory. The flow out
of inventory is a relatively continuous activity where individual items are
placed into the production system for processing. To replenish the
inventory, an order is placed to a supplier. After some delay time, called
the lead time, the raw material is delivered in a lot of a specified amount.
At the moment of delivery, the rate of input is infinite and at other times it
is zero. Whenever the instantaneous rates of input and output to a
component are not the same, the inventory level changes. When the input
rate is higher, inventory grows; when output rate is higher, inventory
declines.
Usually the inventory level remains positive. This corresponds to
the presence of on hand inventory. In cases where the cumulative output
4 Inventory Theory
exceeds the cumulative input, the inventory level is negative. We call this
a backorder or shortage condition. A backorder is a stored output
requirement that is delivered when the inventory finally becomes positive.
Backorders may only be possible for some systems. For example, if the
item is not immediately available the customer may go elsewhere;
alternatively, some items may have an expiration date like an airline seat
and can only be backordered up to the day of departure. In cases where
backorders are impossible, the inventory level is not allowed to become
negative. The demands on the inventory that occur while the inventory
level is zero are called lost sales.
Variability, Uncertainty and Complexity
The are many reasons for variability and uncertainty in inventory systems.
The rates of withdrawal from the system may depend on customer demand
which is variable in time and uncertain in amount. There may be returns
from customers. Lots may be delivered with defects causing uncertainty
in quantities delivered. The lead time associated with an order for
replenishment depends on the capabilities of the supplier which is usually
variable and not known with certainty. The response of a customer to a
shortage condition may be uncertain.
Inventory systems are often complex with one component of the
system feeding another. Figure 3 shows a simple serial manufacturing
system producing a single product.
1
2 3 5
10
Raw
Material
Finished
Goods
4 6 7 9
8
Delay Delay Delay Delay Oper. Oper. Inspect Inspect
Figure 3. A manufacturing system with several locations for inventories
We identify planned inventories in Fig. 3 as inverted triangles,
particularly the raw material and finished goods inventories. Material
passing through the production process is often called work in process
(WIP). These are materials waiting for processing as in the delay blocks
of the figure, materials undergoing processing in the operation blocks, or
materials undergoing inspection in the inspection blocks. All the
components of inventory contribute to the cost of production in terms of
handling and investment costs, and all require management attention.
For our analysis, we will often consider one component of the
system separate from the remainder, particularly the raw material or
finished goods inventories. In reality, rarely can these be managed
independently. The material leaving a raw material inventory does not
leave the system, rather it flows into the remainder of the production
Inventory Models 5
system. Similarly, material entering a finished goods inventory comes
from the system. Any analysis that optimizes one inventory independent
of the others must provide less than an optimal solution for the system as a
whole.
6 Inventory Theory
25.2 The Deterministic Model
An abstraction to the chaotic behavior of Fig. 2 is to assume that items are withdrawn
from the inventory at an even rate a, lots are of a fixed size Q, and lead time is zero or a
constant. The resulting behavior of the inventory is shown in Fig. 4. We use this
deterministic model of the system to explain some of the notation associated with
inventory. Because of its simplicity, we are able to find an optimal solutions to the
deterministic model for several operating assumptions.
s
s+Q
Q/a 2Q/a 3Q/a 4Q/a 5Q/a 6Q/a Time
Inventory Level
0
0
Figure 4. The inventory pattern without uncertainty
Notation
This section lists the factors that are important in making decisions related
to inventories and establishes some of the notation that is used in this
section. Dimensional analysis is sometimes useful for modeling inventory
systems, so we provide the dimensions of each factor. Additional model
dependent notation is introduced later.
• Ordering cost (c(z)): This is the cost of placing an order to an
outside supplier or releasing a production order to a manufacturing
shop. The amount ordered is z and the function c(z) is often
nonlinear. The dimension of ordering cost is ($).
• Setup cost (K): A common assumption is that the ordering cost
consists of a fixed cost, that is independent of the amount ordered,
and a variable cost that depends on the amount ordered. The fixed
cost is called the setup cost and given in ($).
• Product cost (c): This is the unit cost of purchasing the product as
part of an order. If the cost is independent of the amount ordered,
the total cost is cz, where c is the unit cost and z is the amount
ordered. Alternatively, the product cost may be a decreasing
function of the amount ordered. ($/unit)
The Deterministic Model 7
• Holding cost (h): This is the cost of holding an item in inventory
for some given unit of time. It usually includes the lost investment
income caused by having the asset tied up in inventory. This is not
a real cash flow, but it is an important component of the cost of
inventory. If c is the unit cost of the product, this component of the
cost is c , where is the discount or interest rate. The holding cost
may also include the cost of storage, insurance, and other factors
that are proportional to the amount stored in inventory. ($/unit-
time)
• Shortage cost (p): When a customer seeks the product and finds the
inventory empty, the demand can either go unfulfilled or be
satisfied later when the product becomes available. The former
case is called a lost sale, and the latter is called a backorder.
Although lost sales are often important in inventory analysis, they
are not considered in this section, so no notation is assigned to it.
The total backorder cost is assumed to be proportional to the num-
ber of units backordered and the time the customer must wait. The
constant of proportionality is p, the per unit backorder cost per unit
of time. ($/unit-time)
• Demand rate (a): This is the constant rate at which the product is
withdrawn from inventory. (units / time)
• Lot Size (Q): This is the fixed quantity received at each inventory
replenishment. (units)
• Order level (S): The maximum level reached by the inventory is
the order level. When backorders are not allowed, this quantity is
the same as Q. When backorders are allowed, it is less than Q.
(units)
• Cycle time ( ): The time between consecutive inventory
replenishments is the cycle time. For the models of this section =
Q/a. (time)
• Cost per time (T): This is the total of all costs related to the
inventory system that are affected by the decision under
consideration. ($/time)
• Optimal Quantities (Q
*
, S
*
,
*
, T
*
): The quantities defined above
that maximize profit or minimize cost for a given model are the
optimal solution.
Lot Size Model with no Shortages
The assumptions of the model are described in part by Fig. 5, which shows
a plot of inventory level as a function of time. The inventory level ranges
between 0 and the amount Q. The fact that it never goes below 0 indicates
8 Inventory Theory
that no shortages are allowed. Periodically an order is placed for
replenishment of the inventory. The order quantity is Q. The arrival of
the order is assumed to occur instantaneously, causing the inventory level
to shoot from 0 to the amount Q. Between orders the inventory decreases
at a constant rate a. The time between orders is called the cycle time, ,
and is the time required to use up the amount of the order quantity, or Q/a.
Figure 5. Lot size model with no shortages
The total cost expressed per unit time is
Cost/unit time = Setup cost + Product cost + Holding cost
T =
aK
Q
+ ac +
hQ
2
. (3)
In Eq. (3),
a
Q
is the number of orders per unit time. The factor
Q
2
is the
average inventory level. Setting to zero the derivative of T with respect to
Q we obtain
dT
dQ
= –
aK
Q
2
+
h
2
= 0.
Solving for the optimal policy,
Q
*
=
2aK
h
(4)
and
*
=
Q*
a
(5)
Substituting the optimal lot size into the total cost expression, Eq. (3), and
preserving the breakdown between the cost components we see that
T
*
=
ahK
2
+ ac +
ahK
2
= ac + 2ahK (6)
The Deterministic Model 9
At the optimum, the holding cost is equal to the setup cost. We see
that optimal inventory cost is a concave function of product flow through
the inventory (a), indicating that there is an economy of scale associated
with the flow through inventory. For this model, the optimal policy does
not depend on the unit product cost. The optimal lot size increases with
increasing setup cost and flow rate and decreases with increasing holding
cost.
Example 1
A product has a constant demand of 100 units per week. The cost to
place an order for inventory replenishment is $1000. The holding cost for
a unit in inventory is $0.40 per week. No shortages are allowed. Find the
optimal lot size and the corresponding cost of maintaining the inventory.
The optimal lot size from Eq. (4) is
Q
*
=
2(100)(1000)
0.4
= 707.
The total cost of operating the inventory from Eq. (6) is
T
*
= $282.84 per week.
From Q
*
and Eq. (5), we compute the cycle time,
t
*
= 7.07 weeks.
The unit cost of the product was not given in this problem because
it is irrelevant to the determination of the optimal lot size. The product
cost is, therefore, not included in T
*
.
Although these results are easy to apply, a frequent mistake is to
use inconsistent time dimensions for the various factors. Demand may be
measured in units per week, while holding cost may be measured in
dollars per year. The results do not depend on the time dimension that is
used; however, it is necessary that demand be translated to an annual basis
or holding cost translated to a weekly basis.
Shortages Backordered
A deterministic model considered in this section allows shortages to be
backordered. This situation is illustrated in Fig. 6. In this model the
inventory level decreases below the 0 level. This implies that a portion of
the demand is backlogged. The maximum inventory level is S and occurs
when the order arrives. The maximum backorder level is Q – S. A
backorder is represented in the figure by a negative inventory level.
10 Inventory Theory
Figure 6 Lot-size model with shortages allowed
The total cost per unit time is
Cost/time = Setup cost + Product cost + Holding cost + Backorder cost
T =
aK
Q
+ ac +
hS
2
2Q
+
p(Q - S)
2
2Q
(7)
The factor multiplying h in this expression is the average on-hand
inventory level. This is the positive part of the inventory curve shown in
Fig. 6. Because all cycles are the same, the average on-hand inventory
computed for the first cycle is the same as for all time. We see the first
cycle in Fig. 7.
S
0
S-Q
On-Hand
Area
Backorder
Area
Figure 7. The first cycle of the lot size with backorders model
Defining O(t) as the on-hand inventory level and O as the average
on-hand inventory
The Deterministic Model 11
O = (1/ )


0
O(t)dt = (1/ )[On -hand Area]
=
a
Q
¸
¸

_
,

S
2
2a
¸
¸


_
,


=
S
2
2Q

Similarly the factor multiplying p is the average backorder level, B ,
where
B = (1/ )(Backorder Area) =
(Q− S)
2
2Q
.
Setting to zero the partial derivatives of T with respect to Q and S yields
S
*
=
2aK
h

p
p + h
(8)
Q
*
=
2aK
h

p + h
p
(9)
and
*
=
Q
*
a
(10)
Comparing these results to the no shortage case, we see that the optimal
lot size and the cycle times are increased by the factor
[(p + h)/h]
1/2
.
The ratio between the order level and the lot size depends only on the
relative values of holding and backorder cost.
S
*
/Q
*
=
ph
p + h
(11)
This factor is 1/2 when the two costs are equal, indicating that the inven-
tory is in a shortage position one half of the time.
Example 2
We continue Example 1, but now we allow backorders. The backorder
cost is $1 per unit-week. The optimal policy for this situation is found
with Eqs. (8), (9) and (10).
S
*
=
2(100)(1000)
0.4

1
1 + 0.4
= 597.61
Q
*
=
2(100)(1000)
0.4

1 + 0.4
1
= 836.66
12 Inventory Theory
t
*
=
836.66
100
= 8.36 weeks.
Again neglecting the product cost, we find from Eq. (7)
T
*
= $239.04 per week.
The cost of operation has decreased since we have removed the
prohibition against backorders. There backorder level is 239 during each
cycle.
Quantity Discounts
The third deterministic model considered incorporates quantity discount
prices that depend on the amount ordered. For this model no shortages are
allowed, so the inventory pattern appears as in Fig. 5. The discounts will
affect the optimal order quantity. For this model we assume there are N
different prices: c
1
, c
2
, …, c
N
, with the prices decreasing with the index.
The quantity level at which the kth price becomes effective is q
k
, with q
1
equal zero. For purposes of analysis define q
(N+1)
equal to infinity,
indicating that the price c
N
holds for any amount greater than q
N
. Since
the price decreases as quantity increases the values of q
k
increase with the
index k.
To determine the optimal policy for this model we observe that the
optimal order quantity for the no backorder case is not affected by the
product price, c. The value of Q
k
*
would be the same for all price levels if
not for the ranges of order size over which the prices are effective.
Therefore we compute the optimal lot size Q
*
using the parameters of the
problem.
Q* =
2aK
h
. (12)
We then find the optimal order quantity for each price range.
Find for each k the value of Q
k
*
such that
if Q
*
> q
k+1
then Q
k
*
= q
k+1
,
if Q
*
< q
k
then Q
k
*
= q
k
,
if q
k
≤ Q
*
< q
k+1
then Q
k
*
= Q
*
The Deterministic Model 13
Optimal Order Quantity (Q
**
)
a. Find the price level for which Q
*
lies within the quantity range (the
last of the conditions above is true). Let this be level n*. Compute
the total cost for this lot size
T
n*
=
aK
Q*
+ ac
n*
+
hQ*
2
. (13)
b. For each level k > n*, compute the total cost T
k
for the lot size Q
k
*
.
T
k

=
aK
Q
k
*
+ ac
k
+
hQ
k
*
2
(14)
c. Let k
*
be the level that has the smallest value of T
k
. The optimal lot
size Q
**
is the lot size giving the least total cost as calculated in
Steps b and c.
Example 3
We return to the situation of Example 1, but now assume quantity
discounts. The company from which the inventory is purchased hopes to
increase sales by offering a break on the price of the product for larger
orders. For an amount purchased from 0 to 500 units, the unit price is
$100. For orders at or above 500 but less than 1000, the unit price is $90.
This price applies to all units purchased. For orders at or greater than
1000 units, the unit price is $85.
From this data we establish that N = 3. Also
q
1
= 0 and c
1
= 100,
q
2
= 500 and c
2
= 90,
q
3
= 1000 and c
3
= 85,
q
4
= ∞.
Neglecting the quantity ranges, from Eq. (12) we find the optimal lot size
is 707 regardless of price. We observe that this quantity falls in the
second price range. All lower ranges are then excluded. We must then
compare the cost at Q = 707 and c
2
= 90, with the cost at Q = 1000 and c
3
= 85. For the cost c
2
we use Eq. (13).
T
2
= $9282 (for Q
2
*
= 707 and c
2
= 90)
For the cost c
3
we use Eq. (14).
14 Inventory Theory
T
3
= $8,800 (for Q
3
*
= 1000 and c
3
= 85).
Comparing the two costs, we find the optimal policy is to order 1000 for
each replenishment. The cycle time associated with this policy is 10
weeks.
Modeling
The inventory analyst has three principal tasks: constructing the
mathematical model, specifying the values of the model parameters, and
finding the optimal solution. This section has presented only the simplest
cases, with the model specified as the total cost function. The model can
be varied in a number of important aspects. For example, non-
instantaneous replenishment rate, multiple products, and constraints on
maximum inventory are easily incorporated.
When a deterministic model contains a nonlinear total cost
function with only a few variables, the tools of calculus can often be used
find the optimal solution. Some assumptions, however, lead to complex
optimization problems requiring nonlinear programming or other
numerical methods.
The classic lot size formulas derived in this section are based on a
number of assumptions that are usually not satisfied in practice. In
addition it is often difficult to accurately estimate the parameters used in
the formulas. With the admitted difficulties of inaccurate assumptions and
parameter estimation, one might question whether the lot size formulas
should be used at all. We should point out that whether or not the
formulas are used, lot size decisions are frequently required. However
abstract the models are, they do recognize important relationships between
the various cost factors and the lot size, and they do provide answers to lot
sizing questions.
Stochastic Inventory Models 15
25.3 Stochastic Inventory Models
There is no question that uncertainty plays a role in most inventory management
situations. The retail merchant wants enough supply to satisfy customer demands, but
ordering too much increases holding costs and the risk of losses through obsolescence or
spoilage. An order too small increases the risk of lost sales and unsatisfied customers.
The water resources manager must set the amount of water stored in a reservoir at a level
that balances the risk of flooding and the risk of shortages. The operations manager sets
a master production schedule considering the imprecise nature of forecasts of future
demands and the uncertain lead time of the manufacturing process. These situations are
common, and the answers one gets from a deterministic analysis very often are not
satisfactory when uncertainty is present. The decision maker faced with uncertainty does
not act in the same way as the one who operates with perfect knowledge of the future.
In this section we deal with inventory models in which the stochastic nature of
demand is explicitly recognized. Several models are presented that again are only
abstractions of the real world, but whose answers can provide guidance and insight to the
inventory manager.
Probability Distribution for Demand
The one feature of uncertainty considered in this section is the demand for
products from the inventory. We assume that demand is unknown, but
that the probability distribution of demand is known. Mathematical
derivations will determine optimal policies in terms of the distribution.
• Random Variable for Demand (x): This is a random variable that is
the demand for a given period of time. Care must be taken to
recognize the period for which the random variable is defined
because it differs among the models considered.
• Discrete Demand Probability Distribution Function (P(x)): When
demand is assumed to be a discrete random variable, P(x) gives the
probability that the demand equals x.
• Discrete Cumulative Distribution Function (F(b)): The probability
that demand is less than or equal to b is F(b) when demand is
discrete.
F(b) = P(x)
x·0
b

• Continuous Demand Probability Density Function (f(x)): When
demand is assumed to be continuous, f(x) is its density function. The
probability that the demand is between a and b is
P(a ≤ X ≤ b) = f (x)dx
a
b

.
16 Inventory Theory
We assume that demand is nonnegative, so f(x) is zero for negative
values.
• Continuous Cumulative Distribution Function (F(b)): The
probability that demand is less than or equal to b when demand is
continuous.
F(b) = f (x)dx
0
b

• Standard Normal Distribution Function ( (x) and (x)): These are
the density function and cumulative distribution function for the
standard normal distribution.
• Abbreviations: In the following we abbreviate probability
distribution function or probability density function as pdf. We
abbreviate the cumulative distribution function as CDF.
Selecting a Distribution
An important modeling decision concerns which distribution to use for
demand. A common assumption is that individual demand events occur
independently. This assumption leads to the Poisson distribution when the
expected demand in a time interval is small and the normal distribution
when the expected demand is large. Let a be the average demand rate.
Then for an interval of time t the expected demand is at. The Poisson
distribution is then
P(x) =
(at)
x
e
−(at)
x!
.
When at is large the Poisson distribution can be approximated with
a normal distribution with mean and standard deviation
= at , and = at .
Values of F(b) are evaluated using tables for the standard normal
distribution. We include these tables at the end of this chapter.
Of course other distributions can be assumed for demand.
Common assumptions are the normal distribution with other values of the
mean and standard deviation, the uniform distribution, and the exponential
distribution. The latter two are useful for their analytical simplicity.
Finding the Expected Shortage and the Expected Excess
We are often concerned about the relation of demand during some time
period relative to the inventory level at the beginning of the time period.
If the demand is less than the initial inventory level, there is inventory
remaining at the end of the interval. This is the condition of excess. If the
Stochastic Inventory Models 17
demand is greater than the initial inventory level, we have the condition of
shortage.
At some point, assume the inventory level is a positive value z.
During some interval of time, the demand is a random variable x with pdf,
f(x), and CDF, F(x). The mean and standard deviation of this distribution
are and , respectively. With the given distribution, we compute the
probability of a shortage, P
s
, and the probability of excess, P
e
. For a
continuous distribution
P
s
= P{x > z} = f (x)dx
z


= 1 – F(z) (15)
P
e
= P{x ≤ z} = f (x)dx
0
z

= F(z) (16)
In some cases we may also be interested in the expected shortage,
E
s
. This depends on whether the demand is greater or less than z.
Items short =
¹
'
¹
0, if x ≤ z
x – z, if x > z
Then E
s
is the expected shortage and is
E
s
= (x – z) f (x)dx
z


. (17)
Similarly for excess, the expected excess is E
e
E
e
= (z – x) f (x)dx
0
z

The expected excess is expressed in terms of E
s
E
e
= (z – x) f (x)dx
0


– (z – x) f (x)dx
z


= z – µ + E
s
. (18)
For discrete distributions, sums replace the integrals in Eqs. (15)
through (18).
P
s
= P{x ≥ z} = P(x)dx
x· z


= 1 – F(z), (19)
P
e
= P{x ≤ z} = P(x)dx
x·0
z

= F(z). (20)
18 Inventory Theory
E
s
= (x – z)P(x)dx
x· z


. (21)
E
e
= (z – x)P(x)dx
x·0
z

= z – µ + E
s
. (22)
When the Distribution of Demand is Normal
When the demand during the lead time has a normal distribution, tables
are used to find these quantities. Assume the demand during the lead time
has a normal distribution with mean and standard deviation . We
specify the inventory level in terms of the number of standard deviations
away from the mean.
z = + k or k =
z –
We have included at the end of this chapter, a table for the standard
normal distribution, (y), (y) and G(y). We have formerly identified
the first two of these functions as the pdf and CDF. The third is defined as
G(k) · (y − k) (y)dy
k


· (k) − k 1− (k) [ ].
Using the relations between the normal distribution and the standard
normal, the following relationships hold.
f(z) = (1/ ) (k) (23)
F(z) = (k) (24)
E
s
(z) = G(k) (25)
E
e
= z – µ + G(k) (26)
We have occasion to use these results in subsequent examples.
Single Period Stochastic Inventories 19
25.4 Single Period Stochastic Inventories
This section considers an inventory situation in which the current order for the
replenishment of inventory can be evaluated independently of future decisions. Such
cases occur when inventory cannot be added later (spares for a space trip, stocks for the
Christmas season), or when inventory spoils or becomes obsolete (fresh fruit, current
newspapers). The problem may have multiple periods, but the current inventory decision
must be independent of future periods. First we assume there is no setup cost for placing
a replenishment order, and then we assume that there is a setup cost.
Single Period Model with No Setup Cost
Consider an inventory situation where the merchant must purchase a
quantity of items that is offered for sale during a single interval of time.
The items are purchased for a cost c per unit and sold for a price b per
unit. If an item remains unsold at the end of the period, it has a salvage
value of a. If the demand is not satisfied during the interval, there is a cost
of d per unit of shortage. The demand during the period is a random
variable x with given pdf and CDF. The problem is to determine the
number of items to purchase. We call this the order level, S, because the
purchase brings the inventory to level S. For this section, there is no cost
for placing the order for the items.
The expression for the profit during the interval depends on
whether the demand falls above or below S. If the demand is less than S,
revenue is obtained only for the number sold, x, while the quantity
purchased is S. Salvage is obtained for the unsold amount S – x. The
profit in this case is
Profit = bx – cS + a(S – x) for x ≤ S.
If the demand is greater than S, revenue is obtained only for the number
sold, S. A shortage cost of d is expended for each item short, x – S. The
profit in this case is
Profit = bS – cS – d(x – S) for x ≥ S.
Assuming a continuous distribution and taking the expectation over all
values of the random variable, the expected profit is
E[Profit] = b xf (x)dx
0
S

+ b Sf (x)dx
S


– cS + a (S – x) f (x)dx
0
S

– d (x – S) f (x)dx
S


.
Rearranging and simplifying,
20 Inventory Theory
E[Profit] = b – cS + a (S – x) f (x)dx
0
S

– (d + b) (x – S) f (x)dx
S


.
We recognize in this expression the expected excess, E
e
, and the expected
shortage, E
s
. The profit is written in these terms as
E[Profit] = b – cS + aE
e
– (d + b)E
s
(27)
To find the optimal order level, we set the derivative of profit with respect
to S equal to zero.
dE[Profit]
dS
= –c + a f (x)dx
0
S

+ (d + b) f (x)dx
S


= 0.
or –c + aF(S) + (d + b)[1 – F(S)] = 0.
The CDF of the optimal order level, S
*
, is determined by
F(S
*
) =
b – c + d
b – a + d
. (28)
This result is sometimes expressed in terms of the purchasing cost,
c, a holding cost h, expended for every unit held at the end of the period,
and a cost p, expended for every unit of shortage at the end of the period.
In these terms the optimal expected cost is
E[Cost] = cS + hE
e
+ pE
s
.
The optimal solution has
F(S
*
) =
p – c
p + h
. (29)
The two solutions are equivalent if we identify
h = –a = negative of the salvage value
p = b + d = lost revenue per unit + shortage cost.
If the demand during the period has a normal distribution with
mean and standard deviation and , the expected profit is easily
evaluated for any given order level. The order level is expressed in terms
of the number of standard deviations from the mean, or
S = + k .
The optimality condition becomes
Single Period Stochastic Inventories 21
(k
*
) =
b – c + d
b – a + d
=
p – c
p + h
. (30)
The expected value of profit is evaluated with the expression
E[Profit] = b – cS + a[S – + G(k)] – (d + b) G(k). (31)
Call the quantity on the right of the Eq. (28) or (29) the threshold.
Optimality conditions for the order level give values for the CDF. For
continuous random variables there is a solution if the threshold is in the
range from 0 to 1. No reasonable values of the parameters will result in a
threshold less than 0 or larger than 1.
For discrete distributions the optimal value of the order level is the
smallest value of S such that
E[Profit |S + 1] ≤ E[Profit | S + 1].
By manipulation of the summation terms that define the expected profit,
we can show that the optimal order level is the smallest value of S whose
CDF equals or exceeds the threshold. That is
F(S
*
) ≥
b – c + d
b – a + d
or
p – c
p + h
. (32)
Example 4: Newsboy Problem
The classic illustration of this problem involves a newsboy who must
purchase a quantity of newspapers for the day's sale. The purchase cost of
the papers is $0.10 and they are sold to customers for a price of $0.25.
Papers unsold at the end of the day are returned to the publisher for $0.02.
The boy does not like to disappoint his customers (who might turn
elsewhere for supply), so he estimates a "good will" cost of $0.15 for each
customer who is not be satisfied if the supply of papers runs out. The boy
has kept a record of sales and shortages, and estimates that the mean
demand during the day is 250 and the standard deviation is 50. A Normal
distribution is assumed. How many papers should he purchase?
This is a single-period problem because today's newspapers will be
obsolete tomorrow. The factors required by the analysis are
a = 0.02, the salvage value of a newspaper,
b = 0.25, the selling price of each paper,
c = 0.10, the purchase cost of each paper,
d = 0.15, the penalty cost for a shortage.
22 Inventory Theory
Because the demand distribution is normal, we have from Eq. (30),
(k
*
) =
b – c + d
b – a + d
=
0.25 – 0.10 + 0.15
0.25 – 0.02 + 0.15
= 0.7895.
From the normal distribution table, we find that
(0.80) = 0.7881 and (0.85) = 0.8022.
With linear interpolation, we determine k
*
= 0.805. Then
S
*
= (0.805)(50) + 250 = 290.2.
Rounding up, we suggest that the newsboy should purchase 291 papers for
the day. The risk of a shortage during the day is
1 – F(S
*
) = 0.211.
Interpolating in the G(k) column in Table 4, we find that
G(k
*
) = G(0.805) = 0.1192.
Then from Eqs. (25), (26) and (31),
E
e
= 46.2, E
s
= 5.96, and E[Profit] = $32.02 per day.
Example 5: Spares Provisioning
A submarine has a very critical component that has a reliability problem.
The submarine is beginning a 1-year cruise, and the supply officer must
determine how many spares of the component to stock. Analysis shows
that the time between failures of the component is 6 months. A failed
component cannot be repaired but must be replaced from the spares stock.
Only the component actually in operation may fail; components in the
spares stock do not fail. If the stock is exhausted, every additional failure
requires an expensive resupply operation with a cost of $75,000 per
component. The component has a unit cost of $10,000 if stocked at the
beginning of the cruise. Component spares also use up space and other
scarce resources. To reflect these factors a cost of $25,000 is added for
every component remaining unused at the end of the trip. There is
essentially no value to spares remaining at the end of the trip because of
technical obsolescence.
This is a single-period problem because the decision is made only
for the current trip. Failures occur at random, with an average rate of 2
per year. Thus the expected number of failures during the cruise is 2. The
number of failures has a Poisson distribution. The second form of the
solution, Eq. (29), is convenient in this case.
Single Period Stochastic Inventories 23
h = 25,000, the extra cost of storage.
c = 10,000, the purchase cost of each component.
p = 75,000 the cost of resupply.
Expressed in thousands, the threshold is
F(S
*
) =
p – c
p + h
=
7 5 – 1 0
75 + 25
= 0.65.
From the cumulative Poisson distribution using a mean of 2, we find
F(0) = 0.135, F(1) = 0.406, F(2) = 0.677, F(3) = 0.857.
Because this is a discrete distribution, we select the smallest value of S
such that the CDF exceeds 0.65. This occurs for S
*
= 2 which means,
somewhat surprisingly, that only two spares should be brought. This is in
addition to the component initially installed, so that only on the third
failure will a resupply be required. The probability of one or more
resupply operations is
1 – F(2) = 0.323.
The relevance of this model is due in part to the resupply aspect of
the problem. If the system simply stopped after the spares were exhausted
and a single cost of failure were expended, then the assumption of the
linear cost of lost sales would be violated.
Single Period Model with a Fixed Ordering Cost
When the merchant has an initial source of product and there is a fixed
cost for ordering new items, it may be less expensive to purchase no
additional items than to order up to some order level. In this section, we
assume that initially there are z items in stock. If more items are
purchased to increase the stock to a level S, a fixed ordering charge K is
expended. We want to determine a level s, called the reorder point, such
that if z is greater than s we do not purchase additional items. Such a
policy is called the reorder point, order level system, or the (s, S) system.
We consider first the case where additional product is ordered to
bring the inventory to S at the start of the period. The expression for the
expected profit is the same as developed previously, except we must
subtract the ordering charge and it is only necessary to purchase (S – z)
items.
P
O
(z, S) = b – c(S – z) + aE
e
[S] – (d + b)E
s
[S] – K (33)
24 Inventory Theory
We include the argument S with E
e
[S] and E
s
[S] to indicate that these
expected values are computed with the starting inventory level at S.
Neither z nor K affect the optimal solution, and as before
F(S
*
) =
b – c + d
b – a + d

If no addition items are purchased, the system must suffice with
the initial inventory z. The expected profit in this case is
P
N
(z)

=b + aE
e
[z] – (d + b)E
s
[z], (34)
where the expected excess and shortage depend on z.
When z equals S, P
N
is greater than P
O
by the amount K, and
certainly no additional items should be purchased. As z decreases, P
N
and
P
O
become closer. The two expressions are equal when z equals s, the
optimal reorder point. Then the optimal reorder point is s* where,
P
O
(s*, S) = P
N
(s*)
Generally it is difficult to evaluate the integrals that allow this
equation to be solved. When the demand has a normal distribution,
however, the expected profit in the two cases can be written as a function
of the distribution parameters.
Assuming a normal distribution and given the initial supply, z, the
profit when we replenish the inventory up to the level S is
P
O
(z, S) = b – c(S – z) + a[S – + G(k)] – (d + b)[ G(k)] – K (35)
Here S = + k . If we choose not to replenish the inventory, but rather
operate with the items on hand the profit is
P
N
(z) = E[Profit] = b + a[z – + G(k
z
)] – (d + b)[ G(k
z
)]. (31)
Here z = + k
z
.
We modify the newsboy problem by assuming that the boy gets a
free stock of papers each morning. The question is whether he should
order more? The cost of placing an order is $10. In Fig. 8, we have
plotted these the costs with and without an order. The profit is low when
the initial stock is low and we do not reorder. The two curves cross at
about 210. This is the reorder point for the newsboy. If he has 210 papers
or less, he should order enough papers to bring his stock to 291. If he has
more than 210 papers, he should not restock. The profit for a given day
depends on how many papers the boy starts with. The higher of the two
curves in Fig. 8 shows the daily profit if one follows the optimal policy.
As expected the profit grows with the number of free papers.
Single Period Stochastic Inventories 25
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
100 120 140 160 180 200 220 240 260 280 300
Reorder
Not reorder
Prof it
Init ial St ock
Figure 8. Determining the reorder point for the newsboy problem
Example 6: Demand with a Uniform Distribution
The demand for the next period is a random variable with a uniform
distribution ranging from 50 to 250 units. The purchase cost of an item is
$100. The selling price is $150. Items unsold at the end of the period go
"on sale" for $20. All remaining are disposed of at this price. If the
inventory is not sufficient, sales are lost, with a penalty equal to the selling
price of the item. The current level of inventory is 100 units. Additional
items may be ordered at this time; however, a delivery fee will consist of a
fixed charge of $500 plus $10 per item ordered. Should an order be
placed, and if so, how many items should be ordered?
To analyze this problem first determine the parameters of the
model.
c = $110, the purchase cost plus the variable portion of the delivery fee
K = $500, the fixed portion of the delivery fee
p = $150, the lost income associated with a lost sale
h = –$20, the negative of the salvage value of the product.
From Eq. (29), the order level is S, such that
26 Inventory Theory
F(S
*
) =
p – c
p + h
=
150 – 1 1 0
1 5 0 –20
= 0.3077.
Setting the CDF for the uniform distribution equal to this value and
solving for S,
F(S) =
S – 50
250 – 50
= 0.3077 or S = 111.5.
Rounding up, we select S
*
= 112.
`Modifying the expected cost function to include the initial stock
and the cost of placing and order.
C
O

= c(S – z) + hE
e
[S] + pE
s
[S] + K
For the uniform distribution ranging from A to B,
E
e
[S] =
1
(B – A)


A
S
(S – x)dx =
(S – A)
2
2(B – A)

E
s
[S] =
1
(B – A)


S
B
(x – S)dx =
(B – S)
2
2(B – A)

C
O

= c(S – z) + K +
h(S – A)
2
+ p(B – S)
2
2(B – A)

When no order is placed, the purchase cost and the reorder cost terms drop
out and z replaces S.
C
N

=
h(z – A)
2
+ p(B – z)
2
2(B – A)
.
Evaluating C
O

with the order level equal to 112, we find that
C
O
= 19,729 – 110z.
Expressing C
N
entirely in terms of z,
C
N

= –0.05(z – 50)
2
+ 0.375(250 – z)
2
Setting C
O

equal to C
N
, substituting s for z, we solve for the optimal
reorder point.
19729 – 110s = –0.05(s – 50)
2
+ 0.375(250 – s)
2
0.325s
2
– 72.5s + 3543.3 = 0
Single Period Stochastic Inventories 27
Solving the quadratic
1
we find the solutions
s = 150.8 and s = 72.3.
The solution lying above the order level is meaningless, so we select the
reorder point of 72. At this point, for
s = 72.3, we have C
N
= C
O
= 11,814.
Because the current inventory level of 100 falls above the reorder
point, no additional inventory should be purchased. If there were no fixed
charge for delivery, the order would be for 12 units.
Example 7: Demand with an Exponential Distribution
Consider the situation of Example 6 except that demand has an
exponential distribution with a mean value = 150. At the optimal order
level
F(S
*
) = 1 – exp(–S/ ) = 0.3077.
Solving for S, we get
S = – [ln(1 – 0.3077)] = 55.17.
The difference between s and S for the exponential distribution is
approximately
∆ = S – s =
2 K
c + h
=
2(150)(500)
100 − 20
= 41
s = 56 – 41 = 15
For this distribution of demand, the current inventory of 100 is
considerably above both the reorder point and the order level. Certainly
an order should not be placed.

1
The solution to the quadratic ax
2
+ b x + c = 0 is x =
–b ± b
2
–4a c]
2a
.
28 Inventory Theory
25.5 The (s, Q) Inventory Policy
We now consider inventory systems similar to the deterministic models presented in
Section 25.2, but allow the demand to be stochastic. There are a number of ways one
might operate an inventory system with random demand. At this time, we consider the
(s, Q) inventory policy, alternatively called the reorder point, order quantity system.
Figure 9 shows the inventory pattern determined by the (s, Q) inventory policy. The
model assumes that the inventory level is observed at all times. This is called continuous
review. When the level declines to some specified reorder point, s, an order is placed for
a lot size, Q. The order arrives to replenish the inventory after a lead time, L.
Time
Inventory Level
0
0
L L L L L
Q
r
Figure 9. Inventory Operated with the reorder point-lot size Policy
Model
The values of s and Q are the two decisions required to implement the
policy. The lead time is assumed known and constant. The only
uncertainty is associated with demand. In Fig. 9, we show the decrease in
inventory level between replenishments as a straight line, but in reality the
inventory decreases in a stepwise and uneven fashion due to the discrete
and random nature of the demand process.
If we assume that L is relatively small compared to the expected
time required to exhaust the quantity Q, it is likely that only one order is
outstanding at any one time. This is the case illustrated in the figure. We
call the period between sequential order arrivals an order cycle. The cycle
begins with the receipt of the lot, it progresses as demand depletes the
inventory to the level s, and then it continues for the time L when the next
lot is received. As we see in the figure, the inventory level increases
instantaneously by the amount Q with the receipt of an order.
In the following analysis, we are most concerned with the
possibility of shortage during an order cycle, that is the event of the
inventory level falling below zero. This is also called the stockout event.
We assume shortages are backordered and are satisfied when the next
The (s, Q) Inventory Policy 29
replenishment arrives. To determine probabilities of shortages, one need
only be concerned about the random variable that is the demand during the
lead time interval. This is the random variable X with pdf, f(x), and CDF
F(x). The mean and standard deviation of the distribution are and
respectively. The random demand during the lead time gives rise to the
possibility that the inventory level will be depleted before the
replenishment arrives. With the average rate of demand equal to a, the
mean demand during the lead time is
= aL
A shortage will occur if the demand during the period L is greater
that s. This probability, defined as P
s
, is
P
s
= P{x > s} = f (x)dx
s


= 1 – F(s).
The service level is the probability that the inventory will not be depleted
during one order cycle, or
Service level = 1 – P
s
= F(s).
In practical instances the reorder point is significantly greater than
the mean demand during the lead time so that P
s
is quite small. The safety
stock, SS, is defined as
SS = s – .
This is the inventory maintained to protect the system against the
variability of demand. It is the expected inventory level at the end of an
order cycle (just before a replenishment arrives). This is seen in Fig. 10,
where we show the (s, Q) policy for deterministic demand. This figure
will also be useful for the cost analysis of the system.
Time
Inventory Level
0
0
L L
Q
s
L L
SS
Figure 10. The (s, Q) policy for deterministic demand
30 Inventory Theory
General Solution for the (s, Q) Policy
We develop here a general cost model for the (s, Q) policy. The model
and its optimal solution depends on the assumption we make regarding the
cost effects of shortage. The model is approximate in that we do not
explicitly model all the effects of randomness. The principal assumption
is that stockouts are rare, a practical assumption in many instances. In the
model we use the same notation as for the deterministic models of Section
25.2. Since demand is a random variable, we use a as the time averaged
demand rate per unit time.
When we assume that the event of a stockout is rare and inventory
declines in a continuous manner between replenishments, the average
inventory is approximately
Average inventory level =
Q
2
+ s – .
Because the per unit holding cost is h, the holding cost per unit time is
Expected holding cost per unit time = h(
Q
2
+ s – ).
With the backorder assumption, the time between orders is random with a
mean value of Q/a. The cost for replenishment is K, so the expected
replenishment cost per unit time is
Expected replenishment cost per unit time =
Ka
Q
.
With the (s, Q) policy and the assumption that L is relatively
smaller than the time between orders, Q/a, the shortage cost per cycle
depends only on the reorder point. We call this C
s
, and we observe that it
is a function of the reorder point s. We investigate several alternatives for
the definition of this shortage cost. Dividing this cost by the length of a
cycle we obtain
Expected Shortage cost per unit time =
a
Q
C
s
.
Combining these terms we have the general model for the expected cost of
the (s, Q) policy.
EC(s, Q) = h(
Q
2
+ s – ) Inventory cost
+
Ka
Q
Replenishment cost
The (s, Q) Inventory Policy 31
+
a
Q
C
s
Shortage cost (37)
There are two variables in this cost function, Q and s. To find the
optimal policy that minimizes cost, we take the partial derivatives of the
expected cost, Eq. (37), with respect to each variable and set them equal
to zero. First, the partial derivative with respect to Q is
∂EC
∂Q
=
h
2

a(K + C
s
)
Q
2
= 0
or Q
*
=
2a(K + C
s
)
h
(38)
We have a general expression for the optimal lot size that depends on the
cost due to shortages. Taking the partial derivative with respect to the
variable s,
∂EC
∂s
= h +
a
Q

¸

¸
,

_ ∂C
s
∂s
= 0,
or
∂C
s
∂s
= –
hQ
a
(39)
The solution for the optimal reorder point depends on the functional form
of the cost of shortage. We consider four different cases in the remainder
of this section
2
.
Case of a Fixed Cost per Stockout
In this case, there is a cost
1
expended whenever there is the event of a
stockout. This cost is independent of the number of items short, just on
the fact that a stockout has occurred. The expected cost per cycle is
C
s
=
1
P{x

> s} =
1
f (x)dx
s


¸
¸


_
,


. (40)
Now the partial derivative of Eq. (40) with respect to s is
∂C
s
∂s
= –
1
f(s).
Combining Eq. (39) with Eq. (40), we have for the optimal value of s
∂C
s
∂s
= –
1
f(s*) = −
hQ
a
,

2
In this article we follow the development in Peterson and Silver [1979], Chapter 7.
32 Inventory Theory
or f(s*) =
hQ
1
a
, (41)
and C
s
=
1
(1 – F(s*)). (42)
Equation (41) is a condition on the value of the pdf at the optimal reorder
point. If no values of the pdf satisfy this equality, select some minimum
safety level as prescribed by management. The pdf may satisfy this
condition at two different values. It can be shown that the cost function is
minimized when f(x) is decreasing, so for a unimodal pdf, select the
greater of the two solutions.
Equation (41) specifying the optimal s* together with the Eq. (38)
for Q
*
define the optimal control parameters. If one of the parameters are
given at a perhaps not optimal value, these equations yield the optimum
for the other parameter. If both parameters are flexible, a successive
approximation method, as illustrated in Example 13, is used to find values
of Q and s that solve the problem.
Example 8: Optimal reorder point given the order quantity (
1
Given)
The monthly demand for a product has a normal distribution with a mean
of 100 and a standard deviation of 20. We adopt a continuous review
policy in which the order quantity is the average demand for one month.
The interest rate used for time value of money calculations is 12% per
year. The purchase cost of the product is $1000. When it is necessary to
backorder, the cost of paperwork is estimated to be $200, independent of
the number backordered. Holding cost is estimated using the interest cost
of the money invested in a unit of inventory. The lead time for this
situation is 1 week. The fixed order cost is $800. Find the optimal
inventory policy.
We must first adopt a time dimension for those data items related
to time. Here we use 1 month. For this selection,
a = 100 units/month
h = 1000(0.01) = $10/unit-month, the unit cost multiplied by the
interest rate (interest rate is 12%/12 = 1% per month)
1
= $1000, the backorder cost, which is independent in time and
number
K = $800, the order cost.
We must also describe the distribution of demand during the lead time.
For convenience we assume that 1 month has 4 weeks and that the
demands in the weeks are independent and identically distributed normal
variates. With these assumptions the weekly demand has
The (s, Q) Inventory Policy 33
= 100/4 = 25, and
2
= 20
2
/4 = 100 or = 10.
The problem specifies the value of Q as 1 month's demand; thus Q = 100.
Using this value in Eq. (41), we find the associated optimal reorder point.
or f(s*) =
hQ
1
a
=
(10)(100)
(1000)(100)
= 0.01.
The pdf of the standard normal distribution is related to a general normal
distribution as
f(s) = (1/ ) (k) or (k) = f(s)
Then in terms of the standard normal we have
(k
*
) =
hQ
1
a
= (10)(0.01) = 0.1.
We look this up in the standard normal table provided at the end of this
chapter to discover k
*
= ±1.66. Taking the larger of the two possibilities
we find
s* = + (1.66) = 25 + 1.66 (10) = 41.6
or 42 (conservatively rounded up). This is the optimal reorder point for
the given value of Q.
Case of a Charge per Unit Short
In some cases, we may also be interested in the expected number of items
backordered during an order cycle, E
s
. This depends on the demand
during the lead time.
Items backordered =
¹
'
¹
0, if x ≤ s
x – s, if x > s
Therefore, the expected shortage is
E
s
= (x – s) f (x)dx
s


.
For this situation, we assume a cost
2
is expended for every unit
short in a stockout event. The expected cost per cycle is
C
s
=
2
E
s
.
Now the partial derivative with respect to s is
∂C
s
∂s
= –
2
f (x)dx
0
s


¸
¸


_
,



= –
2
(1 – F(s)).
34 Inventory Theory
From Eq. (41), the optimal value of s must satisfy
∂C
s
∂s
= –
2
(1 – F(s*)) = −
hQ
a
or F(s*) = 1 –
hQ
2
a
. (43)
In this case, we have a condition on the CDF at the optimal reorder point.
If the expression on the right is less than zero, use some minimum reorder
point specified by management.
For a given value of s, the optimal order quantity is determined
from Eq. (38) by substituting the value of C
s
.
C
s
=
2
E
s
=
2
(x – s*) f (x)dx
s


¸
¸


_
,


. (44)
This integral is difficult to compute except for simple distributions. It is
evaluated with tables for the normal random variable using Eq. (25).
Managers may find it difficult to specify the shortage cost
2
. It is
easier to specify that the inventory meet some service level. One might
require that the inventory meet demands from stock in 99% of the
inventory cycles. The service level is actually the value of F(s). Given
values of h, Q and a, one can compute with Eq. (43) the implied shortage
cost for the given service level.
Example 9: Optimal reorder point given the order quantity (
2
Given)
We consider again Example 8, but change the cost structure for
backorders. Now we assume that we must treat each backordered
customer separately. The cost of paperwork and good will is estimated to
be $200 per unit backordered. This is
2
. The optimal policy is governed
by Eq. (43).
F(s*) = 1 –
hQ
2
a
= 1 –
(10)(100)
(200)(100)
= 0.95.
We know that the probabilities for a normal distribution is related to the
standard normal distribution by
F(s) =
s – ¸
¸

_
,
.
(k
*
) = 0.95.
From the normal distribution table we find that this is associated with a
standard normal variate of z = 1.64. The reorder point is then
The (s, Q) Inventory Policy 35
s* = + (1.64) = 25 + 1.64 (10) = 41.4
or 42 (conservatively rounded up). This is the optimal for the given value
of Q.
Case of a Charge per Unit Short per Unit Time
When the backorder cost depends not only on the number of backorders
but the time a backorder must wait for delivery, we would like to compute
the expected unit-time of backorders for an inventory cycle. When the
number of backorders is x – s and the average demand rate is a, the
average time a customer must wait for delivery is
x – s
2a
.
The resulting unit-time measure for backorders is
(x – s)
2
2a
.
Integrating we find the expected value T
s
, where
T
s
=
1
2a
(x – s)
2
f (x)dx
s


. (45)
We consider here the case when a cost
3
is expended for every
unit short per unit of time. The expected cost per cycle is
C
s
=
3
T
s
. (46)
Now the partial derivative of C
s
with respect to s is
∂C
s
∂s
= –
3
a
(x – s) f (x)dx
s


¸
¸


_
,


= –
3
E
s
a
.
From Eq. (41), the optimal value of s must satisfy
∂C
s
∂s
= −
3
E
s
a
= –
hQ
a
or E
s
(s*) =
hQ
3
. (47)
We have added the parameter s* to the expected shortage to indicate its
value is a function of the reorder point. Note that Silver et al. [1998]
report the result E
s
(s*) =
Qh
h +
3
which is derived using a more accurate
36 Inventory Theory
representation of the average inventory. The two results are
approximately the same when
3
>> h, as assumed here.
Example 10: Optimal reorder point given the order quantity (
3
Given)
We consider again Example 8, but now we assume that $1000 is expended
per unit backorder per month. This is
3
. The optimal policy is governed
by Eq. (47).
E
s
(s*) =
hQ
3
=
(10)(100)
1000
= 1.
When the demand is governed by the normal distribution, the expected
shortage at the optimum is
E
s
(s*) = G(k
*
) = 1
where k
*
=
s * −
or G(k
*
) = 0.1
From the table at the end of the chapter
k
*
= 0.9.
The reorder point is then
s* = + (0.9) = 25 + 9 = 34
This is the optimum for the given value of Q.
Lost Sales Case
In this case sales are not backordered. A customer that arrives when there
is no inventory on hand leaves without satisfaction, and the sale is lost.
When stock is exhausted during the lead time, the inventory level rises to
the level Q when it is finally replenished. The effect of this situation is to
raise the average inventory level by the expected number of shortages in a
cycle, E
s
. We also experience a shortage cost based on the number of
shortages in a stockout event. We use
L
to indicate the cost for each lost
sale. For the case of lost sales the approximate expected cost is
EC(Q, s) = h
Q
2
+ s − + E
s
¸
¸

_
,
Inventory cost
+
aK
Q
Replenishment cost
The (s, Q) Inventory Policy 37
+
a
L
Q
E
s
Shortage cost (48)
Here we are neglecting the fact that with lost sales, not all the demand is
met. The number of orders per unit time is slightly less than a/Q. Taking
partial derivatives with respect to Q and s we find the optimal lot size is
Q
*
=
2a(K +
L
E
s
)
h
(49)
∂EC
∂s
= h 1 +
∂E
s
∂s
¸
¸

_
,
)+
a
L
Q
∂E
s
∂s
¸
¸

_
,
= 0,
or
∂E
s
∂s
= −
hQ
hQ+
L
a
or (1 – F(s*)) =
hQ
hQ+
L
a
F(s*) = 1 –
hQ
hQ+
L
a
=
L
a
hQ+
L
a
(50)
Example 11: Optimal reorder point given the order quantity (
L
Given)
We consider Example 8 again, but now we assume that the sale is lost
given a stockout. We charge $2000 for every lost sale. This is
L
. The
optimal policy is governed by Eq. (50).
F(s*) =
L
a
hQ+
L
a
=
(2000)(100)
(10)(100) + (2000)(100)
= 0.995
From the table at the end of the chapter
k
*
= 2.58.
The reorder point is then
s* = + (2.58) = 25 + 22.6 = 47.6
This is the optimum for the given value of Q.
Summary
We have found in this section solutions for several assumptions regarding
the costs due to shortages. These are summarized below for easy use. The
optimal reorder point requires one to find the value s* that corresponds to
38 Inventory Theory
f(s*), F(s*) or E
s
(s*) equaling some simple function of the problem
parameters.
The optimal order quantity for each case depends on the shortage
cost, C
s
, and is given by
Q* =
2a(K + C
s
)
h
This equation is used directly when a value of s is specified. It is used
iteratively when the optimum for both s and Q is required.
Table 1. The (s, Q) Policy for Continuous Distributions
Situation C
s
Optimal reorder point Normal solution
Fixed cost per
stockout (
1
)
1
(1 – F(s))
f(s*) =
hQ
1
a
(k
*
) =
hQ
1
a
Charge per unit
Short (
2
)
2
E
s
F(s*) = 1 –
hQ
2
a
(k
*
) = 1 –
hQ
2
a
Charge per unit
short per unit time
(
3
)
3
T
s
E
s
(s*) =
hQ
3
G(k
*
) =
hQ
3
Charge per unit of
lost sales (
L
)
L
E
s
F(s*) =
L
a
hQ+
L
a
(k
*
) =
L
a
hQ+
L
a
Determination of the Order Quantity
Up until this point, all our examples have determined the reorder point
given the order quantity. The following examples illustrate the
determination of the order quantity when the reorder point is given, and
the determination of optimal values for both variables simultaneously.
Example 12: Optimal order quantity given the reorder point
We continue from Example 9 in which the shortage cost is
2

= $200 per
unit short. The demand during the lead time is normal with µ = 25 and
= 10. If the reorder point is fixed at 50, what is the optimal order
quantity?
For a normal distribution the expected shortage cost is
C
s
=
2
G(k
s
)
The (s, Q) Inventory Policy 39
where k
s
= (s – )/ . For s = 50 and k
s
= 2.5, G(2.5) = 0.0020, E
s
=
0.020, C
s
= 4. Then the optimal order quantity is
Q
*
=
2a(K + C
s
)
h
=
2(100)(800 + 4)
10
= 126.8
or 127 (conservatively rounded up).
Example 13: Both optimal order quantity and reorder point
In the previous examples we fixed one of the decisions and found the
optimal value of the other. We need an iterative procedure to find both,
Q
*
and s*. We use the expression below sequentially.
Q =
2a(K + C
s
)
h
, (k
s
) = 1 –
hQ
2
a
, C
s
=
2
G(k
s
)
The first step is to assume C
s
= 0 and to find the corresponding
optimal order quantity.
Q
1
= 126.5.
Using this value of Q
1
, we find the optimal reorder point
k
s
= 1.53 or s = 40.3
The expected shortage per period with this reorder point is
C
s
=
2
G(1.53) = (200)(10)(0.02736) = 54.72
For this value of C
s
, we have
Q
2
= 130.7.
Using this value of Q
2
, we find the optimal reorder point
k
s
= 1.51 or s = 40.1
Computing the associated C
s
we find
Q
3
= 130.9.
It appears that the values are converging, so we adopt the policy
Q
*
= 131 and s* = 40.
40 Inventory Theory
25.6 Variations on the (s, Q) Model
Reorder Point Based on Inventory Position
In the foregoing, we have assumed that a replenishment order is to be
placed whenever the inventory level reaches the reorder point. A move
practical idea is to use the inventory position rather than the inventory
level as an indicator. The inventory position is the inventory level plus the
quantity on order. The difference is illustrated in Fig. 11. We note that
inventory level is the same as inventory position when there are no
outstanding orders. In the early cycles of the figure, the inventory level
crosses the reorder point at the same time as the inventory position, and
the same order pattern is obtained using either measure. Basing the order
on the inventory level fails, however, when there is a lead time demand
larger than the order quantity, as in the last cycle of the figure. In this case
the inventory level falls below the reorder point and never reaches it again.
Using the inventory position, however, allows two orders to be placed in
quick succession, thus keeping the inventory in control.
Inventory Position
0
0
L L L L
Q
s
Inventory Level
L
Time
Figure 11. Using inventory position as a measure for placing orders
Using the inventory position in this manner, also allows us to drop
the requirement that the lot size be very much greater than the average
demand during the lead time. The results in the table can be used even in
cases where the lot size is small in relation to the lead time demand. The
primary assumption for the derivations is that the probability of a stockout
be small. This probability depends on the reorder point and not the lot
size.
Variations on the (s, Q) Model 41
When the lot size is small, there may be many outstanding orders
at any given time, emphasizing the need to track the inventory position. A
particularly interesting case is when the lot size is 1. This implies that a
replenishment order is placed whenever an item is withdrawn from
inventory.
Discrete Demand During the Lead Time
The results of the table were derived for continuous distributions. In fact
the items in an inventory are usually discrete, and a discrete demand
distribution may be more appropriate. This is particularly true when the
reorder point is relatively small. For the discrete distribution p(x) is the
probability that the random demand during the lead time takes the value x.
F(x) is the probability that the demand is less than or equal to x. The
expected shortage and expected unit-time shortage are
E
s
= (x – s) p(x)
x ·s +1


(51)
T
s
=
1
2a
(x – s)
2
p(x)
x ·s +1


(52)
Table 2. The (s, Q) Policy for Discrete Distributions
Situation C
s
Optimal reorder point
Fixed cost per
stockout (
1
)
1
(1 – F(s))
p(s* – 1) >
hQ
1
a
≥ p(s*)
Charge per unit
short (
2
)
2
E
s
F(s*) ≤ 1 –
hQ
2
a
< F(s*+1)
Charge per unit
short per unit time
(
3
)
3
T
s
E
s
(s*–1) >
hQ
3
≥ E
s
(s*)
Charge per unit of
lost sales (
L
)
L
E
s
F(s*) ≤ 1 –
hQ
hQ+
L
a
< F(s*+1)
Lead Time a Random Variable
Previously we assumed that lead time is a constant. Indeed this is a very
desirable characteristic of an inventory system. The lead time may
actually be uncertain in duration due to variability in shipping times,
material availability and supplier processing times.
42 Inventory Theory
Let lead time be a random variable Y with pdf h(y), and let demand
be the random variable X with joint pdf g(x, y) -- that is, the demand
distribution depends on the lead time. The marginal distribution of
demand during the lead time is
f(x) = g(x, y)h(y)dy
0


(53)
This pdf can then be used in conjunction with the standard normal
distribution table to determine approximate solutions.
The (R, S) Inventory Policy 43
25.7 The (R, S) Inventory Policy
A different way to manage a stochastic inventory system is illustrated in Fig. 12 This is
called a periodic review policy in that the inventory level is only observed at intervals of
R. If the inventory is at level y, a quantity S – y is ordered to bring the inventory position
to S. S is called the order level. After a lead time interval L, the replenishment order is
delivered. Figure 12 shows the inventory position with dotted lines and the inventory
level with solid lines.
Time
0
0
L
L L L
R R R
S
Figure 12. The (R, S) inventory policy
The analysis of this policy is much like that for the (s, Q) policy. For the
(s, Q) policy, the reorder point s is set to protect against the possibility of
shortage during the lead time L. For the (R, S) policy, the order level S is
set to protect against a shortage in the time interval R + L. In the event of
a particular order at time t, the lowest inventory that is affected by that
order occurs at time t + R + L. The quantity S must be large enough to
keep the probability of a shortage in that time interval small. The (R, S)
policy is much more affected by variability than the (s, Q) policy because
of the longer interval. The advantage of the policy is that it does not
require continuous review.
To analyze this system, we define the demand in the interval R + L
to be the random variable X. The pdf and CDF are f
P
(x) and F
P
(x),
respectively. The mean and variance during the interval are
P
and
P
(the subscript “P” stands for periodic). The cost (per unit time) of
operation of the inventory system expressed in terms of R and S is
EC(R, S) = h (
aR
2
+ S – ) Inventory cost
44 Inventory Theory
+
K
R
Replenishment cost
+
1
R
C
S
Shortage cost (54)
Evaluation of the shortage cost depends on the assumption of the costs
experienced in the event of a stockout.
Fixed cost per stockout: C
s
=
1
(1 – F(S)) =
1
f
P
(x)
S


dx
¸
¸


_
,


(55)
Charge per unit short: C
s
=
2
E
s

=
2
(x – s) f
P
(x
S


)dx
¸
¸


_
,


(56)
Charge per unit short per unit time:
C
s
=
3
T
s
=
1
2a
(x – s)
2
f
P
(x)
S


dx
¸
¸


_
,


(57)
Charge per lost sale: C
s
=
L
E
s

=
L
(x – s) f
P
(x
S


)dx
¸
¸


_
,


(58)
We cannot determine the optimal R easily in this case because C
s
depends on R. In particular the distribution of demand in the expressions
for C
s
have a mean and variance that depends on the interval L + R. If we
assume that the review interval R is set elsewhere, the determination of the
optimal order level follows exactly the derivation for s* for the (s, Q)
policy. The optimal conditions are found in Table 1. We create Table 3
from Table 1 by replacing
s with S, Q with aR, and L with R + L.
A similar table for discrete distributions can be constructed starting from
Table 2.
The (R, S) Inventory Policy 45
Table 3. The (R, S) Policy for Continuous Distributions
Situation C
s
Optimal reorder point Normal solution
Fixed cost per
stockout (
1
)
1
(1 – F(S))
f
P
(S
*
) =
hR
1
(k
*
) =
P
hR
1
Charge per unit
Short (
2
)
2
E
s
F
P
(S
*
) = 1 –
hR
2
(k
*
) = 1 –
hR
2
Charge per unit
short per snit time
(
3
)
3
T
s
E
s
(S
*
) =
haR
3
G(k
*
) =
haR
P 3
Charge per unit of
lost sales (
L
)
L
E
s
F(S
*
) = 1 –
hR
hR +
L
(k
*
) = 1 –
hR
hR +
L
Example 14
We consider again the situation of Example 9 except the inventory is
reviewed every month. The monthly demand for the product has a normal
distribution with a mean of 100 and a standard deviation of 20. The
holding cost is $10 per unit per month. When it is necessary to backorder
a customer request, the cost of paperwork and good will is estimated to be
$200 per unit. The lead time for orders is zero. Find the optimal
inventory policy.
The parameters of the model are as follows
R = 1 month
h = $10, the cost of holding one unit for one month
2
= 200, the cost of backordering one unit
= 100 and = 20 during the review period (the distribution is
normal)
With these values we can determine the optimal policy.
(k
*
) = 1 –
hR
2
= 1 – (10/200) = 0.95
Using the standard normal table, we find that the CDF has the value 0.957
for z = 1.65. The order level is
S = 100 + 20(1.65) = 133.
46 Inventory Theory
The optimal policy is to order a quantity (133 – x) each month. The
probability of a shortage during the month is
1 – (k
*
) = 0.05.
The safety stock is S – µ = 33 units.
Now we consider the same problem when the review period is
every 2 months. The situation is the same except
R = 2 months
= 200 and = 20 2 = 28.28 during the review period (here we
assume that the monthly demands are independent)
(k
*
) = 1 –
hR
2
= 1 – (10)(2)/200) = 0.90.
Using a standard normal table, we find that the CDF has the value 0.899
for k = 1.28. Thus the reorder level is
S = 200 + 28.28(1.28) = 236.
The optimal policy is to order a quantity (236 – x) every two months. The
safety stock is: S – µ = 36. Three extra units of inventory are necessary
for the longer review period. For the continuous review of Example 9 the
safety stock is 17 units.
Exercises 47
25.8 Exercises
Section 25.2
1. It is now January 1. A hardware distributor is reviewing his inventory policy for
hammers, which have a relatively constant demand of 2000 units per month. The
distributor buys the hammers from his supplier for $5 each and sells them for $10.
Every time he places an order for replenishment a shipping and paper preparation
cost of $500 is charged. The distributor's holding cost is 15% (annual) of his
average investment in inventory. (The holding cost is h = (0.15)(5) = $0.75 per
hammer per year.)
Answer each of the following independently ( one part does not depend on another).
a. The current inventory policy is to replenish the inventory every month. What is
the total annual cost of this policy?
b. What is the optimal lot size when no shortages are allowed?
c. The distributor wants to order only at the beginning of the month, so the cycle
time must be in whole months. What is the optimal lot size in this case?
d. An OR analyst suggests that money might be saved if the customers would
accept backorders. If the distributor offers a discount of 1% of the purchase
price for every day the customer has to wait for delivery, what is the optimal
policy?
e. The supplier, in an effort to get larger orders, offers a $0.10 discount on the price
of hammers for an order of at least 10,000 units. Should the distributor take
advantage of this deal? Assume no shortages are allowed.
f. It is now January 1, and the current inventory is 3500 units. There is a lead time
(time between when the order is placed and when it is delivered) of 1 month.
When should the next order be placed for the solutions of parts (b), (c), (d), and
(e) of this problem?
g. Compute the annual profit on hammers for parts (a), (b), (c), (d), and (e).
2. Rather than an instantaneous replenishment of inventory as assumed in Section 25.2,
the inventory is replaced by a production process that can add units to the inventory
at a constant rate r (r > a ). Of course, the process cannot be operated continually
but must be started and stopped. Each time the process is begun a setup cost K is ex-
pended. Otherwise the inventory problem is the same as previously stated. Derive a
formula for the optimal lot size when no shortages are allowed.
3. A meat market advertises that its hamburger is the freshest in town. To get rid of old
stock, the price of the hamburger is reduced by $0.10 per pound for every hour it
remains unsold after it is ground (for simplicity assume this reduction is linear in
time). The demand for hamburger is 250 pounds per hour. No shortages are al-
48 Inventory Theory
lowed. The cost to set up and run the hamburger grinder is $100, independent of the
amount processed. How much should be ground at each setup?
4. Consider a gift shop operator with a relatively constant demand for a certain type of
wall hanging. She can order a case that holds 12 units for a cost of $1200 or a gross
that holds 144 units for $13,000. These costs include delivery and product cost. The
demand for the product is 12 units per month, so the alternatives represent ordering a
1-month or 1-year supply. As an approximation assume that demand is at a constant
rate and no shortages are allowed. A principal concern for this business woman is
the cost of capital, which is 20% per year. Which plan should she adopt?
5. Light bulbs come in cases of 144 bulbs. A case of bulbs costs $200. An office
building uses bulbs at an average rate of 1000 bulbs per month. The company that
sells the bulbs charges $50 per delivery, regardless of the number of cases delivered.
The management of the building uses a 15% annual carrying cost rate for inventory.
No shortages are allowed.
a. Assuming the usage rate for bulbs is constant, how many cases of bulbs should
be ordered in each delivery? Only whole cases can be ordered.
b. With this order quantity, how often will bulbs be delivered?
c. A long-standing policy has been to receive 10 cases per delivery. What is the
annual cost penalty of this nonoptimal policy over the cost of the optimal policy?
6. Find the optimal policy for 5 if backorders are allowed. A failed bulb is simply not
replaced until the beginning of the next month if the supply is exhausted. The
management assumes a "loss of good will" charge of $10 per bulb per month for
backorders.
Section 25.4
7. A garden nursery has a short selling season of 20 days for a certain kind of tree. The
nursery currently has 20 such trees in stock and has one last opportunity to order
more. The expected demand for trees is 3 per day, but the actual demand has a
Poisson distribution. The nursery purchases the trees for $75 and sells them for
$175. Trees remaining at the end of the season must be cared for until the next year.
The cost of such care per tree is $20 and the holding cost on investment is $2. If
demand occurs after the inventory is exhausted, the sales are lost. Find the number
of trees that should be ordered and the probability that this inventory level will be
sufficient to meet all demand. (Note that since leftover trees will actually be kept
after the season and eventually sold, the purchase cost is not relevant. Only the
shortage cost and holding cost should be used.)
8. For the parameters given in Example 6 in the text, the probability of a shortage is
almost 70%. Management has determined that this is too high. A level of 10% is
considered an acceptable probability of shortage. If all other parameters remain the
same, what penalty is implicitly being charged for a shortage? For this value find
Exercises 49
the optimal policy assuming the distribution of demand is uniform between 50 and
250 items.
9. Use the penalty found in Exercise 8, and find the optimal policy if the demand is
exponential distributed with a mean of 150.
10. A dishonest developer wants to make quick money building a condominium. He
sets a 1-year time table for his activities, after which he will leave town for places
unknown. His problem is to determine how many units to build. He estimates the
following probabilities for the demand for his units. In the table below, x is the
number of units demanded and p (x) is the probability of that demand.
x 1 2 3 4 5 6 7 8
p (x) 0.02 0.03 0.09 0.14 0.19 0.14 0.10 0.05
x 9 10 11 12 13 14 15
p (x) 0.05 0.05 0.04 0.04 0.03 0.02 0.01
The selling price for a condo unit is $100,000, while it will cost only $40,000 to
build it. If the developer runs out of units while there is still demand, he charges a
lost opportunity cost of $10,000. If the year passes and he has not sold all his units,
he will have to dump them for $20,000 each. How many units should he build to
maximize his expected profit?
11. A military commander faces a dangerous mission which requires the use of
helicopters. The commander wants to determine how many to bring on the mission.
The helicopters cost $1 million each. At least five are required for mission success.
Any number less than five is judged a serious detriment to the mission. In order to
quantify the situation, the commander has attached a cost of $100 million for every
unit less than five that finishes the mission. Any units that remain at the end of the
mission are destroyed.
Failure-causing events are assumed to occur at random on the average of once every
hour. The number of events is independent of the number of helicopters active.
Each event will destroy one helicopter. The mission lasts 10 hours.
How many helicopters should the commander bring?
12. Consider a problem with the following parameters.
c = $10, h = $1, p = $100, and K = $200.
Find the optimal (s , S) policy when:
a. Demand has a uniform distribution with A = 20 and B = 60
b. Demand has an exponential distribution with a mean value of 40
50 Inventory Theory
Section 25.5 - 25.7
13. We repeat the conditions of Example 9. The demand per month has a normal
distribution with mean 100 and standard deviation 20. The lead time is one week.
Assume four weeks per month.
a = 1200 units/year.
h = $120/unit-year.
2
= $200.
K = $800.
Find the optimal (s, Q) policy when the following changes are made. The changes
are not cumulative.
a. The average demand doubles but the standard deviation remains the same.
b. The standard deviation of demand per week doubles.
c. The fixed setup cost doubles.
d. The holding cost doubles.
e. The backorder cost doubles.
f. The lead time doubles.
14. Find the optimal order level for an inventory system for which orders are placed
every 2 weeks. The weekly demand has a normal distribution with a mean of 25 and
a standard deviation 5. The cost of the item is $20. Interest per year is 20%.
Assume a 50-week year. The backorder cost is $5 per unit.
15. Use the data of Exercise 14 and add the information that the setup cost is $200 and
the lead time is 1 week. Find the optimal continuous review policy for the following
situations.
a. The order quantity is set to the average demand for 1 month.
b. The reorder point is set at the average demand over the lead time.
c. Neither the reorder point nor the order quantity is specified.
16. A supplier of rebuilt engines expects an average of three engine customers every 2
months. Customer arrivals follow a Poisson process. The engines are obtained from
a manufacturer who delivers 2 months after an order is placed.
a. What should be the supplier's reorder point if she wants a 60% chance of not
having a shortage during the lead time?
b. What should be the reorder point if the lead time were 1 month?
Exercises 51
17. Consider the light bulb situation of Exercise 5, except that demand is stochastic
rather than deterministic. The weekly demand for light bulbs is a random variable
with a normal distribution that has a mean of 250 bulbs and a standard deviation of
50 bulbs. An order for bulbs is placed once a month (on the first of the month) and
the order is delivered right away. If the supply of bulbs runs out and a bulb fails, the
tenant simply must wait until the beginning of the next month. For purposes of
analysis, this event is assumed to have a cost of $10. Assume for simplicity that
months are 4 weeks long and that there are 48 weeks in a year. What specific
numerical rule should the management use to determine how much to order each
month?
18. The average demand for a product at a warehouse is 1200 units per year. Customers
arrive at random (a Poisson process) so the exact demand for any given period of
time cannot be computed. The warehouse manager replenishes the inventory by a
monthly order to the factory. The size of the order equals the previous month's
sales.
a. What kind of inventory system is this?
b. What order level will provide a 95% chance that the inventory will not run out
during an order cycle?
19. The manager in Exercise 18 installs computerized inventory control so that
continuous review is possible. He adopts a (s, Q) policy. A lead time of 1 week
passes between when an order is placed and when the inventory is replenished. A
lot size of 100 is selected. Assume a month is 4 weeks.
a. What reorder point will provide a 95% chance that the inventory will not run out
during an order cycle?
b. What is the average cycle time for this plan?
20. A gasoline distributor has a weekly demand that is approximately normally
distributed. The average demand is 10,000 gallons per week; however, the standard
deviation is 3000 gallons. His supply is replenished every 6 weeks. He must pay
$0.75 per gallon. His cost of capital is 0.5% per week. He recovers this cost in the
price he charges for gasoline, but if any remains in inventory when the next order
arrives he charges a holding cost based on this interest rate and the value of the
inventory. If the distributor runs out of supply during the period, he "borrows" gas
from other distributors, paying an extra $0.10 per gallon for the privilege. The
borrowed gas must be returned when his supply is replenished. What is the optimal
inventory policy for the distributor?
21. One of the disadvantages of using large order quantities in a production process is,
of course, large holding costs. Another disadvantage is the inflexibility associated
with having a large inventory. If a design change is to be incorporated into the
52 Inventory Theory
product or some custom feature is to be added for particular customers, the
modification must await the next production cycle while the entire inventory is
depleted. The average length of the cycle is Q /a, so as Q grows so does the length
of the cycle. This might be called the lead time of the production process. The
reason for a large order quantity is a large setup cost, so reducing setup cost results
in reduced holding cost, decreased production lead time, and increased flexibility.
This is one of the tenants of the "just-in-time" production systems, which go to great
measures to reduce setup costs. Analyze the following production systems that are
characterized by different setup costs K, and different annual costs to implement the
production process B.
a. K = $10,000 , B = $10,000
b. K = $5,000 , B = $30,000
c. K = $1,000 , B = $110,000
In each case weekly demand follows a normal distribution with mean 50, and
standard deviation 10. Use a 50-week year and an (s, Q) inventory policy. The
holding cost is h = $200/unit-year. The lead time when a reorder is placed is 1
week. The shortage cost is $100 per unit. Compute the optimal values of s and Q
for each case and compare the total costs (including B) and production lead times.
Choose the least cost system.
Exercises 53
Table 4. Function Values for the Standard Normal Distribution, y ∈ −3,0 [ ]

y ( y) ( y) G( y) y ( y) ( y) G( y)
-3.00 0.0044 0.0013 3.0004 -1.50 0.1295 0.0668 1.5293
-2.95 0.0051 0.0016 2.9505 -1.45 0.1394 0.0735 1.4828
-2.90 0.0060 0.0019 2.9005 -1.40 0.1497 0.0808 1.4367
-2.85 0.0069 0.0022 2.8506 -1.35 0.1604 0.0885 1.3909
-2.80 0.0079 0.0026 2.8008 -1.30 0.1714 0.0968 1.3455
-2.75 0.0091 0.0030 2.7509 -1.25 0.1826 0.1056 1.3006
-2.70 0.0104 0.0035 2.7011 -1.20 0.1942 0.1151 1.2561
-2.65 0.0119 0.0040 2.6512 -1.15 0.2059 0.1251 1.2121
-2.60 0.0136 0.0047 2.6015 -1.10 0.2179 0.1357 1.1686
-2.55 0.0154 0.0054 2.5517 -1.05 0.2299 0.1469 1.1257
-2.50 0.0175 0.0062 2.5020 -1.00 0.2420 0.1587 1.0833
-2.45 0.0198 0.0071 2.4523 -0.95 0.2541 0.1711 1.0416
-2.40 0.0224 0.0082 2.4027 -0.90 0.2661 0.1841 1.0004
-2.35 0.0252 0.0094 2.3532 -0.85 0.2780 0.1977 0.9600
-2.30 0.0283 0.0107 2.3037 -0.80 0.2897 0.2119 0.9202
-2.25 0.0317 0.0122 2.2542 -0.75 0.3011 0.2266 0.8812
-2.20 0.0355 0.0139 2.2049 -0.70 0.3123 0.2420 0.8429
-2.15 0.0396 0.0158 2.1556 -0.65 0.3230 0.2578 0.8054
-2.10 0.0440 0.0179 2.1065 -0.60 0.3332 0.2743 0.7687
-2.05 0.0488 0.0202 2.0574 -0.55 0.3429 0.2912 0.7328
-2.00 0.0540 0.0228 2.0085 -0.50 0.3521 0.3085 0.6978
-1.95 0.0596 0.0256 1.9597 -0.45 0.3605 0.3264 0.6637
-1.90 0.0656 0.0287 1.9111 -0.40 0.3683 0.3446 0.6304
-1.85 0.0721 0.0322 1.8626 -0.35 0.3752 0.3632 0.5981
-1.80 0.0790 0.0359 1.8143 -0.30 0.3814 0.3821 0.5668
-1.75 0.0863 0.0401 1.7662 -0.25 0.3867 0.4013 0.5363
-1.70 0.0940 0.0446 1.7183 -0.20 0.3910 0.4207 0.5069
-1.65 0.1023 0.0495 1.6706 -0.15 0.3945 0.4404 0.4784
-1.60 0.1109 0.0548 1.6232 -0.10 0.3970 0.4602 0.4509
-1.55 0.1200 0.0606 1.5761 -0.05 0.3984 0.4801 0.4244
-1.50 0.1295 0.0668 1.5293 0.00 0.3989 0.5000 0.3989

y is a standard normal variate
(y) is the probability density function, pdf
(y) is the cumulative distribution function, CDF
G(y) = (y) – y[1 – (y)]
54 Inventory Theory
Table 4. (Cont.) Function Values for the Standard Normal Distribution y ∈ 0,3 [ ]

y ( y) ( y) G( y) y ( y) ( y) G( y)
0.00 0.3989 0.5000 0.3989 1.50 0.1295 0.9332 0.0293
0.05 0.3984 0.5199 0.3744 1.55 0.1200 0.9394 0.0261
0.10 0.3970 0.5398 0.3509 1.60 0.1109 0.9452 0.0232
0.15 0.3945 0.5596 0.3284 1.65 0.1023 0.9505 0.0206
0.20 0.3910 0.5793 0.3069 1.70 0.0940 0.9554 0.0183
0.25 0.3867 0.5987 0.2863 1.75 0.0863 0.9599 0.0162
0.30 0.3814 0.6179 0.2668 1.80 0.0790 0.9641 0.0143
0.35 0.3752 0.6368 0.2481 1.85 0.0721 0.9678 0.0126
0.40 0.3683 0.6554 0.2304 1.90 0.0656 0.9713 0.0111
0.45 0.3605 0.6736 0.2137 1.95 0.0596 0.9744 0.0097
0.50 0.3521 0.6915 0.1978 2.00 0.0540 0.9772 0.0085
0.55 0.3429 0.7088 0.1828 2.05 0.0488 0.9798 0.0074
0.60 0.3332 0.7257 0.1687 2.10 0.0440 0.9821 0.0065
0.65 0.3230 0.7422 0.1554 2.15 0.0396 0.9842 0.0056
0.70 0.3123 0.7580 0.1429 2.20 0.0355 0.9861 0.0049
0.75 0.3011 0.7734 0.1312 2.25 0.0317 0.9878 0.0042
0.80 0.2897 0.7881 0.1202 2.30 0.0283 0.9893 0.0037
0.85 0.2780 0.8023 0.1100 2.35 0.0252 0.9906 0.0032
0.90 0.2661 0.8159 0.1004 2.40 0.0224 0.9918 0.0027
0.95 0.2541 0.8289 0.0916 2.45 0.0198 0.9929 0.0023
1.00 0.2420 0.8413 0.0833 2.50 0.0175 0.9938 0.0020
1.05 0.2299 0.8531 0.0757 2.55 0.0154 0.9946 0.0017
1.10 0.2179 0.8643 0.0686 2.60 0.0136 0.9953 0.0015
1.15 0.2059 0.8749 0.0621 2.65 0.0119 0.9960 0.0012
1.20 0.1942 0.8849 0.0561 2.70 0.0104 0.9965 0.0011
1.25 0.1826 0.8944 0.0506 2.75 0.0091 0.9970 0.0009
1.30 0.1714 0.9032 0.0455 2.80 0.0079 0.9974 0.0008
1.35 0.1604 0.9115 0.0409 2.85 0.0069 0.9978 0.0006
1.40 0.1497 0.9192 0.0367 2.90 0.0060 0.9981 0.0005
1.45 0.1394 0.9265 0.0328 2.95 0.0051 0.9984 0.0005
1.50 0.1295 0.9332 0.0293 3.00 0.0044 0.9987 0.0004
Bibliography 55
Bibliography
Askin, R.G. and C.R. Standridge, Modeling and Analysis of Manufacturing Systems, John
Wiley & Sons, New York, 1993.
Elsayed, E.A. and T.O. Boucher, Analysis and Control of Production Systems, Second
Edition, Prentice Hall, Englewood Cliffs, NJ, 1994.
Hopp, W. and M. Spearman, Factory Physics, Second Edition, McGraw Hill, New York,
2000.
Johnson, L.A. and D.C. Montgomery, Operations Research in Production Planning,
Scheduling, and Inventory Control, John Wiley & Sons, New York, 1974.
Moskowitz, H. and G.P. Wright, Operations Research Techniques for Management,
Prentice Hall, Englewood Cliffs, NJ, 1979.
Nahmis, S., Production and Operations Analysis, Third Edition, Irwin, Chicago, 1997.
Peterson, R. and E.A. Silver, Decision Systems for Inventory Management and
Production Planning, John Wiley & Sons, New York, 1979.
Ritzman, L.P. and L.J. Krajewski, Foundations of Operations Management, Prentice
Hall, Upper Saddle River, NJ, 2003.
Russell, R.S. and B.W. Taylor III, Operations Management, Fourth Edition, Prentice
Hall, Upper Saddle River, NJ, 2003.
Silver, E.A., D.F. Pyke and R. Peterson, Inventory Management and Production
Planning and Scheduling, Third Edition, John Wiley & Sons, New York, 1998.
Sipper, D., Production: Planning, Control and Integration, McGraw Hill, New York,
1997.
Tersine, R.J., Principles of Inventory and Materials Management, Fourth Edition,
Prentice Hall, Prentice Hall, Englewood Cliffs, NJ, 1994.
Zipkin, P.H., Foundations of Inventory Management, McGraw Hill, New York, 2000.

2

Inventory Theory

regarding the costs of operation. Sections 25.5 and 25.6 derive optimal solutions for the (s, S) policy under a variety of conditions. This policy places an order up to level S when the inventory level falls to the reorder point s. Section 25.7 extends these results to the (R, S) policy. In this case, the inventory is observed periodically (with a time interval R), and is replenished to level S. Flow, Inventory and Time An inventory is represented in the simple diagram of Fig. 1. Items flow into the system, remain for a time and then flow out. Inventories occur whenever the time an individual enters is different than when it leaves. During the intervening interval the item is part of the inventory.

Flow In

Inventory Level (Residence Time)

Flow Out

Figure 1. A system component with inventory For example, say the box in Fig. 1 represents a manufacturing process that takes a fixed amount of time. A product entering the box at one moment leaves the box one hour later. Products arrive at a rate of 100 per hour. Clearly, if we look in the box, we will find some number of items. That number is the inventory level. The relation between flow, time and inventory level that is basic to all systems is Inventory level = (Flow rate )(Residence time) where the flow rate is expressed in the same time units as the residence time. For the example, we have Inventory Level = (100 products/hour )(1 hour) = 100 products. When the factors in Eq. (1) are not constant in time, we typically use their mean values. Whenever two of the factors in the above expression are given, the third is easily computed. Consider a queueing system for which customers are observed to arrive at an average rate of 10 per hour. When the customer finds the servers busy, he or she must wait. Customers in the system, either waiting or be served, are the inventory for this system. Using a sampling procedure we determine that the average number of customers in the inventory is 5. We ask, how long on the average is each customer in the system? Using the relation between the flow, time and (1)

Inventory Models inventory, we determine the answer as 0.5 hours. As we saw in the Chapter 16, Queueing Models, Eq (1) is called Little's Law.

3

The relation between time and inventory is significant, because very often reducing the throughput time for a system is just as important as reducing the inventory level. Since they are proportional, changing one factor inevitably changes the other. The Inventory Level The inventory level depends on the relative rates of flow in and out of the system. Define y(t) as the rate of input flow at time t and Y(t) the cumulative flow into the system. Define z(t) as the rate of output flow at time t and Z(t) as the cumulative flow out of the system. The inventory level, I(t) is the cumulative input less the cumulative output. ⌡ ⌡ I(t) = Y(t) – Z(t) = ⌠y(x)dx - ⌠z(x)dx
0 0 t t

(2)

Figure 2 represents the inventory for a system when the rates vary with time.
Inventory Level

0 0

Time

Figure 2. Inventory fluctuations as a function of time The figure might represent a raw material inventory. The flow out of inventory is a relatively continuous activity where individual items are placed into the production system for processing. To replenish the inventory, an order is placed to a supplier. After some delay time, called the lead time, the raw material is delivered in a lot of a specified amount. At the moment of delivery, the rate of input is infinite and at other times it is zero. Whenever the instantaneous rates of input and output to a component are not the same, the inventory level changes. When the input rate is higher, inventory grows; when output rate is higher, inventory declines. Usually the inventory level remains positive. This corresponds to the presence of on hand inventory. In cases where the cumulative output

4

Inventory Theory exceeds the cumulative input, the inventory level is negative. We call this a backorder or shortage condition. A backorder is a stored output requirement that is delivered when the inventory finally becomes positive. Backorders may only be possible for some systems. For example, if the item is not immediately available the customer may go elsewhere; alternatively, some items may have an expiration date like an airline seat and can only be backordered up to the day of departure. In cases where backorders are impossible, the inventory level is not allowed to become negative. The demands on the inventory that occur while the inventory level is zero are called lost sales.

Variability, Uncertainty and Complexity The are many reasons for variability and uncertainty in inventory systems. The rates of withdrawal from the system may depend on customer demand which is variable in time and uncertain in amount. There may be returns from customers. Lots may be delivered with defects causing uncertainty in quantities delivered. The lead time associated with an order for replenishment depends on the capabilities of the supplier which is usually variable and not known with certainty. The response of a customer to a shortage condition may be uncertain. Inventory systems are often complex with one component of the system feeding another. Figure 3 shows a simple serial manufacturing system producing a single product.
1 2 3 Oper. 4 Delay 5 Inspect 6 Delay 7 Oper. 8 Delay 9 10

Raw Delay Material

Inspect Finished Goods

Figure 3. A manufacturing system with several locations for inventories We identify planned inventories in Fig. 3 as inverted triangles, particularly the raw material and finished goods inventories. Material passing through the production process is often called work in process (WIP). These are materials waiting for processing as in the delay blocks of the figure, materials undergoing processing in the operation blocks, or materials undergoing inspection in the inspection blocks. All the components of inventory contribute to the cost of production in terms of handling and investment costs, and all require management attention. For our analysis, we will often consider one component of the system separate from the remainder, particularly the raw material or finished goods inventories. In reality, rarely can these be managed independently. The material leaving a raw material inventory does not leave the system, rather it flows into the remainder of the production

Inventory Models

5

system. Similarly, material entering a finished goods inventory comes from the system. Any analysis that optimizes one inventory independent of the others must provide less than an optimal solution for the system as a whole.

so we provide the dimensions of each factor. The dimension of ordering cost is ($). The fixed cost is called the setup cost and given in ($). Dimensional analysis is sometimes useful for modeling inventory systems. and a variable cost that depends on the amount ordered. lots are of a fixed size Q. Additional model dependent notation is introduced later. • Setup cost (K): A common assumption is that the ordering cost consists of a fixed cost. We use this deterministic model of the system to explain some of the notation associated with inventory. the total cost is cz. the product cost may be a decreasing function of the amount ordered. The resulting behavior of the inventory is shown in Fig.2 The Deterministic Model An abstraction to the chaotic behavior of Fig. Inventory Level s+Q s 0 0 Q/a 2Q/a 3Q/a 4Q/a 5Q/a 6Q/a Time Figure 4. 2 is to assume that items are withdrawn from the inventory at an even rate a. Alternatively. The amount ordered is z and the function c(z) is often nonlinear. 4. • Product cost (c): This is the unit cost of purchasing the product as part of an order. If the cost is independent of the amount ordered.6 Inventory Theory 25. we are able to find an optimal solutions to the deterministic model for several operating assumptions. Because of its simplicity. The inventory pattern without uncertainty Notation This section lists the factors that are important in making decisions related to inventories and establishes some of the notation that is used in this section. ($/unit) . that is independent of the amount ordered. and lead time is zero or a constant. • Ordering cost (c(z)): This is the cost of placing an order to an outside supplier or releasing a production order to a manufacturing shop. where c is the unit cost and z is the amount ordered.

which shows a plot of inventory level as a function of time. ($/time) • Optimal Quantities (Q*.The Deterministic Model 7 • Holding cost (h): This is the cost of holding an item in inventory for some given unit of time. ($/unittime) • Shortage cost (p): When a customer seeks the product and finds the inventory empty. S*. The fact that it never goes below 0 indicates . Although lost sales are often important in inventory analysis. (units) • Cycle time ( ): The time between consecutive inventory replenishments is the cycle time. this component of the cost is c . The holding cost may also include the cost of storage. T*): The quantities defined above that maximize profit or minimize cost for a given model are the optimal solution. insurance. If c is the unit cost of the product. and other factors that are proportional to the amount stored in inventory. The constant of proportionality is p. The total backorder cost is assumed to be proportional to the number of units backordered and the time the customer must wait. it is less than Q. The former case is called a lost sale. (time) • Cost per time (T): This is the total of all costs related to the inventory system that are affected by the decision under consideration. the demand can either go unfulfilled or be satisfied later when the product becomes available. they are not considered in this section. and the latter is called a backorder. (units) • Order level (S): The maximum level reached by the inventory is the order level. When backorders are allowed. but it is an important component of the cost of inventory. where is the discount or interest rate. Lot Size Model with no Shortages The assumptions of the model are described in part by Fig. (units / time) • Lot Size (Q): This is the fixed quantity received at each inventory replenishment. It usually includes the lost investment income caused by having the asset tied up in inventory. the per unit backorder cost per unit of time. The inventory level ranges between 0 and the amount Q. For the models of this section = Q/a. ($/unit-time) • Demand rate (a): This is the constant rate at which the product is withdrawn from inventory. When backorders are not allowed. this quantity is the same as Q. so no notation is assigned to it. 5. *. This is not a real cash flow.

The order quantity is Q. Solving for the optimal policy. Q* = 2aK h (4) (5) Q* and * = a Substituting the optimal lot size into the total cost expression. causing the inventory level to shoot from 0 to the amount Q. Figure 5. . (3). The factor 2 is the average inventory level. Q is the number of orders per unit time. (3) a Q In Eq. Lot size model with no shortages The total cost expressed per unit time is Cost/unit time = Setup cost + Product cost + Holding cost aK hQ T = Q + ac + 2 . Eq.8 Inventory Theory that no shortages are allowed. (3). Setting to zero the derivative of T with respect to Q we obtain dT aK h dQ = – Q2 + 2 = 0. The arrival of the order is assumed to occur instantaneously. The time between orders is called the cycle time. Periodically an order is placed for replenishment of the inventory. and is the time required to use up the amount of the order quantity. and preserving the breakdown between the cost components we see that T* = ahK 2 + ac + ahK 2 = ac + 2ahK (6) . or Q/a. Between orders the inventory decreases at a constant rate a.

The results do not depend on the time dimension that is used. not included in T*. while holding cost may be measured in dollars per year. indicating that there is an economy of scale associated with the flow through inventory. The maximum backorder level is Q – S. it is necessary that demand be translated to an annual basis or holding cost translated to a weekly basis. A backorder is represented in the figure by a negative inventory level. Find the optimal lot size and the corresponding cost of maintaining the inventory. (6) is T* = $282.84 per week. The optimal lot size from Eq.07 weeks. t* = 7. therefore. 6. Although these results are easy to apply. (4) is Q* = 2(100)(1000) = 707. The maximum inventory level is S and occurs when the order arrives. The product cost is. the holding cost is equal to the setup cost. Shortages Backordered A deterministic model considered in this section allows shortages to be backordered. For this model.4 The total cost of operating the inventory from Eq. The unit cost of the product was not given in this problem because it is irrelevant to the determination of the optimal lot size. however. we compute the cycle time.40 per week. No shortages are allowed. The cost to place an order for inventory replenishment is $1000. The holding cost for a unit in inventory is $0. This situation is illustrated in Fig. (5). Demand may be measured in units per week. 0. the optimal policy does not depend on the unit product cost.The Deterministic Model 9 At the optimum. In this model the inventory level decreases below the 0 level. This implies that a portion of the demand is backlogged. . Example 1 A product has a constant demand of 100 units per week. The optimal lot size increases with increasing setup cost and flow rate and decreases with increasing holding cost. From Q* and Eq. a frequent mistake is to use inconsistent time dimensions for the various factors. We see that optimal inventory cost is a concave function of product flow through the inventory (a).

S)2 T = Q + ac + 2Q + 2Q (7) The factor multiplying h in this expression is the average on-hand inventory level. 7.10 Inventory Theory Figure 6 Lot-size model with shortages allowed The total cost per unit time is Cost/time = Setup cost + Product cost + Holding cost + Backorder cost aK hS2 p(Q . This is the positive part of the inventory curve shown in Fig. We see the first cycle in Fig. S On-Hand Area Backorder Area 0 S-Q Figure 7. The first cycle of the lot size with backorders model Defining O(t) as the on-hand inventory level and O as the average on-hand inventory . Because all cycles are the same. 6. the average on-hand inventory computed for the first cycle is the same as for all time.

The optimal policy for this situation is found with Eqs. Example 2 We continue Example 1. (8). 2Q Setting to zero the partial derivatives of T with respect to Q and S yields S* = Q* = 2aK h 2aK h p p+h p+h p (8) (9) (10) Q* and * = a Comparing these results to the no shortage case.4 = 836.The Deterministic Model 11 ⌡ O = (1/ )⌠O(t)dt = (1/ )[On -hand Area] 0  a   S2  S2 =     = 2Q    Q  2a  Similarly the factor multiplying p is the average backorder level. ph S*/Q*= p + h (11) This factor is 1/2 when the two costs are equal. The backorder cost is $1 per unit-week. indicating that the inventory is in a shortage position one half of the time. but now we allow backorders. we see that the optimal lot size and the cycle times are increased by the factor [(p + h)/h]1/2.4 1 1 + 0. where B = (1/ )(Backorder Area) = (Q − S)2 .4 2(100)(1000) 0. B . The ratio between the order level and the lot size depends only on the relative values of holding and backorder cost.61 1 + 0. (9) and (10).4 = 597.66 1 . S* = Q* = 2(100)(1000) 0.

so the inventory pattern appears as in Fig. with q1 equal zero. There backorder level is 239 during each cycle.04 per week. Quantity Discounts The third deterministic model considered incorporates quantity discount prices that depend on the amount ordered.36 weeks. (12) We then find the optimal order quantity for each price range. The cost of operation has decreased since we have removed the prohibition against backorders.12 Inventory Theory 836. For this model no shortages are allowed. if qk ≤ Q* < qk+1 then Qk* = Q* . 5. (7) T* = $239. if Q* < qk then Qk* = qk.66 t* = 100 = 8. …. For this model we assume there are N different prices: c1. indicating that the price cN holds for any amount greater than qN. with the prices decreasing with the index. The quantity level at which the kth price becomes effective is qk. The value of Qk* would be the same for all price levels if not for the ranges of order size over which the prices are effective. Find for each k the value of Qk* such that if Q* > qk+1 then Qk* = qk+1. The discounts will affect the optimal order quantity. c. c2. we find from Eq. For purposes of analysis define q(N+1) equal to infinity. Therefore we compute the optimal lot size Q* using the parameters of the problem. To determine the optimal policy for this model we observe that the optimal order quantity for the no backorder case is not affected by the product price. Again neglecting the product cost. Since the price decreases as quantity increases the values of qk increase with the index k. cN. Q* = 2aK h .

but now assume quantity discounts. the unit price is $90. from Eq. We observe that this quantity falls in the second price range. (13). q2 = 500 and c2 = 90. This price applies to all units purchased. (13) b. * T2 = $9282 (for Q2 = 707 and c2 = 90) (14) c. Also q1 = 0 and c1 = 100. Example 3 We return to the situation of Example 1. Let this be level n*. Let k* be the level that has the smallest value of Tk. Find the price level for which Q* lies within the quantity range (the last of the conditions above is true). Neglecting the quantity ranges. The company from which the inventory is purchased hopes to increase sales by offering a break on the price of the product for larger orders. hQk* aK Tk = Q * + ack + 2 k size Q** is the lot size giving the least total cost as calculated in Steps b and c. All lower ranges are then excluded. For an amount purchased from 0 to 500 units. (14). For the cost c2 we use Eq. . We must then compare the cost at Q = 707 and c2 = 90. Compute the total cost for this lot size aK hQ* Tn* = Q* + acn* + 2 . For each level k > n*. compute the total cost Tk for the lot size Qk* .The Deterministic Model Optimal Order Quantity (Q** ) 13 a. From this data we establish that N = 3. The optimal lot For the cost c3 we use Eq. For orders at or greater than 1000 units. the unit price is $85. (12) we find the optimal lot size is 707 regardless of price. For orders at or above 500 but less than 1000. with the cost at Q = 1000 and c3 = 85. the unit price is $100. q3 = 1000 and c3 = 85. q4 = ∞.

Some assumptions. In addition it is often difficult to accurately estimate the parameters used in the formulas.800 (for Q3 = 1000 and c3 = 85). We should point out that whether or not the formulas are used. However abstract the models are. with the model specified as the total cost function. however. For example. Comparing the two costs. and constraints on maximum inventory are easily incorporated. Modeling The inventory analyst has three principal tasks: constructing the mathematical model. The model can be varied in a number of important aspects. one might question whether the lot size formulas should be used at all. we find the optimal policy is to order 1000 for each replenishment. The cycle time associated with this policy is 10 weeks. noninstantaneous replenishment rate. This section has presented only the simplest cases.14 Inventory Theory * T3 = $8. specifying the values of the model parameters. lead to complex optimization problems requiring nonlinear programming or other numerical methods. multiple products. When a deterministic model contains a nonlinear total cost function with only a few variables. The classic lot size formulas derived in this section are based on a number of assumptions that are usually not satisfied in practice. With the admitted difficulties of inaccurate assumptions and parameter estimation. lot size decisions are frequently required. and finding the optimal solution. the tools of calculus can often be used find the optimal solution. they do recognize important relationships between the various cost factors and the lot size. . and they do provide answers to lot sizing questions.

In this section we deal with inventory models in which the stochastic nature of demand is explicitly recognized.Stochastic Inventory Models 15 25.3 Stochastic Inventory Models There is no question that uncertainty plays a role in most inventory management situations. These situations are common. • Random Variable for Demand (x): This is a random variable that is the demand for a given period of time. An order too small increases the risk of lost sales and unsatisfied customers. but whose answers can provide guidance and insight to the inventory manager. The decision maker faced with uncertainty does not act in the same way as the one who operates with perfect knowledge of the future. • Discrete Demand Probability Distribution Function (P(x)): When demand is assumed to be a discrete random variable. Care must be taken to recognize the period for which the random variable is defined because it differs among the models considered. • Discrete Cumulative Distribution Function (F(b)): The probability that demand is less than or equal to b is F(b) when demand is discrete. f(x) is its density function. The retail merchant wants enough supply to satisfy customer demands. The operations manager sets a master production schedule considering the imprecise nature of forecasts of future demands and the uncertain lead time of the manufacturing process. F(b) = ∑ P(x) x =0 b • Continuous Demand Probability Density Function (f(x)): When demand is assumed to be continuous. The water resources manager must set the amount of water stored in a reservoir at a level that balances the risk of flooding and the risk of shortages. and the answers one gets from a deterministic analysis very often are not satisfactory when uncertainty is present. but ordering too much increases holding costs and the risk of losses through obsolescence or spoilage. P(x) gives the probability that the demand equals x. The probability that the demand is between a and b is P(a ≤ X ≤ b) = ∫ f (x)dx . Probability Distribution for Demand The one feature of uncertainty considered in this section is the demand for products from the inventory. Mathematical derivations will determine optimal policies in terms of the distribution. a b . but that the probability distribution of demand is known. Several models are presented that again are only abstractions of the real world. We assume that demand is unknown.

Then for an interval of time t the expected demand is at. If the . Selecting a Distribution An important modeling decision concerns which distribution to use for demand. Values of F(b) are evaluated using tables for the standard normal distribution. Of course other distributions can be assumed for demand. there is inventory remaining at the end of the interval. If the demand is less than the initial inventory level. Common assumptions are the normal distribution with other values of the mean and standard deviation. This assumption leads to the Poisson distribution when the expected demand in a time interval is small and the normal distribution when the expected demand is large. • Abbreviations: In the following we abbreviate probability distribution function or probability density function as pdf. We abbreviate the cumulative distribution function as CDF. Finding the Expected Shortage and the Expected Excess We are often concerned about the relation of demand during some time period relative to the inventory level at the beginning of the time period. x! When at is large the Poisson distribution can be approximated with a normal distribution with mean and standard deviation = at . • Continuous Cumulative Distribution Function (F(b)): The probability that demand is less than or equal to b when demand is continuous. A common assumption is that individual demand events occur independently.16 Inventory Theory We assume that demand is nonnegative. The latter two are useful for their analytical simplicity. the uniform distribution. This is the condition of excess. so f(x) is zero for negative values. We include these tables at the end of this chapter. and = at . The Poisson distribution is then P(x) = (at) x e −(at) . F(b) = ∫ f (x)dx 0 b • Standard Normal Distribution Function ( (x) and (x)): These are the density function and cumulative distribution function for the standard normal distribution. and the exponential distribution. Let a be the average demand rate.

sums replace the integrals in Eqs. the demand is a random variable x with pdf. assume the inventory level is a positive value z. and CDF. Ps = P{x ≥ z} = ∑ P(x)dx = 1 – F(z). Ps. x =0 z . Es. and the probability of excess. The mean and standard deviation of this distribution are and . (15) through (18). Pe. F(x). This depends on whether the demand is greater or less than z. (18) For discrete distributions.Stochastic Inventory Models 17 demand is greater than the initial inventory level. we compute the probability of a shortage. During some interval of time. f(x). x=z ∞ (19) (20) Pe = P{x ≤ z} = ∑ P(x)dx = F(z). we have the condition of shortage. With the given distribution. the expected excess is Ee Ee = ∫ (z – x) f (x)dx 0 z The expected excess is expressed in terms of Es Ee = ∫ (z – x) f (x)dx – 0 ∞ ∫ (z – x) f (x)dx z ∞ = z – µ + Es. z ∞ (17) Similarly for excess. At some point. 0. For a continuous distribution Ps = P{x > z} = ∫ f (x)dx = 1 – F(z) z ∞ (15) Pe = P{x ≤ z} = ∫ f (x)dx = F(z) 0 z (16) In some cases we may also be interested in the expected shortage. if x ≤ z Items short = x – z. if x > z  Then Es is the expected shortage and is Es = ∫ (x – z) f (x)dx . respectively.

(y) and G(y). z= +k or k = z– We have included at the end of this chapter. the following relationships hold. (23) (24) (25) (26) . We have formerly identified the first two of these functions as the pdf and CDF. The third is defined as G(k) = ∞ ∫ (y − k) k (y)dy = (k) − k [1− (k)] . (y). x =0 (22) When the Distribution of Demand is Normal When the demand during the lead time has a normal distribution. tables are used to find these quantities. We specify the inventory level in terms of the number of standard deviations away from the mean. Assume the demand during the lead time has a normal distribution with mean and standard deviation .18 Inventory Theory Es = z ∑ (x – z)P(x)dx . f(z) = (1/ ) (k) F(z) = (k) Es(z) = G(k) Ee = z – µ + G(k) We have occasion to use these results in subsequent examples. x=z ∞ (21) Ee = ∑ (z – x)P(x)dx = z – µ + Es. Using the relations between the normal distribution and the standard normal. a table for the standard normal distribution.

First we assume there is no setup cost for placing a replenishment order. The profit in this case is Profit = bS – cS – d(x – S) for x ≥ S.Single Period Stochastic Inventories 19 25.4 Single Period Stochastic Inventories This section considers an inventory situation in which the current order for the replenishment of inventory can be evaluated independently of future decisions. Single Period Model with No Setup Cost Consider an inventory situation where the merchant must purchase a quantity of items that is offered for sale during a single interval of time. x – S. x. . but the current inventory decision must be independent of future periods. The profit in this case is Profit = bx – cS + a(S – x) for x ≤ S. If the demand is greater than S. If the demand is not satisfied during the interval. or when inventory spoils or becomes obsolete (fresh fruit. stocks for the Christmas season). The items are purchased for a cost c per unit and sold for a price b per unit. The demand during the period is a random variable x with given pdf and CDF. The expression for the profit during the interval depends on whether the demand falls above or below S. the expected profit is E[Profit] = b ∫ xf (x)dx + b ∫ Sf (x)dx 0 S S ∞ – cS + a ∫ (S – x) f (x)dx – d ∫ (x – S) f (x)dx . revenue is obtained only for the number sold. Assuming a continuous distribution and taking the expectation over all values of the random variable. Salvage is obtained for the unsold amount S – x. Such cases occur when inventory cannot be added later (spares for a space trip. S. For this section. If an item remains unsold at the end of the period. We call this the order level. because the purchase brings the inventory to level S. there is no cost for placing the order for the items. while the quantity purchased is S. current newspapers). 0 S S ∞ Rearranging and simplifying. it has a salvage value of a. S. there is a cost of d per unit of shortage. The problem may have multiple periods. A shortage cost of d is expended for each item short. If the demand is less than S. revenue is obtained only for the number sold. and then we assume that there is a setup cost. The problem is to determine the number of items to purchase.

is determined by b–c+d F(S*) = b – a + d . expended for every unit held at the end of the period. a holding cost h. c. The two solutions are equivalent if we identify h = –a = negative of the salvage value p = b + d = lost revenue per unit + shortage cost. dS 0 S or –c + aF(S) + (d + b)[1 – F(S)] = 0. or S= The optimality condition becomes +k . expended for every unit of shortage at the end of the period. the expected profit is easily evaluated for any given order level. (28) This result is sometimes expressed in terms of the purchasing cost. The profit is written in these terms as E[Profit] = b – cS + aEe – (d + b)Es (27) To find the optimal order level. The order level is expressed in terms of the number of standard deviations from the mean. and the expected shortage. S*. 0 S S ∞ E[Profit] = b We recognize in this expression the expected excess. (29) . and a cost p. If the demand during the period has a normal distribution with mean and standard deviation and . The optimal solution has p–c F(S*) = p + h . Ee. S ∞ The CDF of the optimal order level.20 Inventory Theory – cS + a ∫ (S – x) f (x)dx – (d + b) ∫ (x – S) f (x)dx . dE[Profit] = –c + a ∫ f (x)dx + (d + b) ∫ f (x)dx = 0. we set the derivative of profit with respect to S equal to zero. Es. In these terms the optimal expected cost is E[Cost] = cS + hEe+ pEs.

the salvage value of a newspaper. (32) . That is b–c+d p–c F(S*) ≥ b – a + d or p + h .02. The boy has kept a record of sales and shortages.15 for each customer who is not be satisfied if the supply of papers runs out.15. No reasonable values of the parameters will result in a threshold less than 0 or larger than 1. the penalty cost for a shortage.Single Period Stochastic Inventories b–c+d p–c (k*) = b – a + d = p + h .10. b = 0. (28) or (29) the threshold. Papers unsold at the end of the day are returned to the publisher for $0.02. By manipulation of the summation terms that define the expected profit.25.10 and they are sold to customers for a price of $0.25. For discrete distributions the optimal value of the order level is the smallest value of S such that E[Profit |S + 1] ≤ E[Profit | S + 1]. The factors required by the analysis are a = 0. The expected value of profit is evaluated with the expression E[Profit] = b – cS + a[S – + G(k)] – (d + b) G(k). Example 4: Newsboy Problem The classic illustration of this problem involves a newsboy who must purchase a quantity of newspapers for the day's sale. we can show that the optimal order level is the smallest value of S whose CDF equals or exceeds the threshold. The boy does not like to disappoint his customers (who might turn elsewhere for supply). the selling price of each paper. and estimates that the mean demand during the day is 250 and the standard deviation is 50. Optimality conditions for the order level give values for the CDF. so he estimates a "good will" cost of $0. How many papers should he purchase? This is a single-period problem because today's newspapers will be obsolete tomorrow. d = 0. the purchase cost of each paper. The purchase cost of the papers is $0. c = 0. For continuous random variables there is a solution if the threshold is in the range from 0 to 1. 21 (30) (31) Call the quantity on the right of the Eq. A Normal distribution is assumed.

is convenient in this case.7895.000 is added for every component remaining unused at the end of the trip.85) = 0. components in the spares stock do not fail. every additional failure requires an expensive resupply operation with a cost of $75. Failures occur at random.15 = 0.1192. Thus the expected number of failures during the cruise is 2.15 (k*) = b – a + d = 0. and the supply officer must determine how many spares of the component to stock. Example 5: Spares Provisioning A submarine has a very critical component that has a reliability problem. and E[Profit] = $32.000 if stocked at the beginning of the cruise. Interpolating in the G(k) column in Table 4. The second form of the solution. (30). Eq. The submarine is beginning a 1-year cruise. With linear interpolation. we find that (0. Then S* = (0.10 + 0. A failed component cannot be repaired but must be replaced from the spares stock. If the stock is exhausted. . Only the component actually in operation may fail. Ee = 46.211. Component spares also use up space and other scarce resources. To reflect these factors a cost of $25. Rounding up.2. we find that G(k*) = G(0.805)(50) + 250 = 290. (26) and (31). we have from Eq. The number of failures has a Poisson distribution.22 Inventory Theory Because the demand distribution is normal. Then from Eqs. Analysis shows that the time between failures of the component is 6 months. we determine k* = 0. we suggest that the newsboy should purchase 291 papers for the day.7881 and (0. b–c+d 0. The component has a unit cost of $10.8022.2.25 – 0.80) = 0. with an average rate of 2 per year. Es = 5.02 + 0.805. From the normal distribution table. (29). There is essentially no value to spares remaining at the end of the trip because of technical obsolescence. (25). This is a single-period problem because the decision is made only for the current trip.000 per component. The risk of a shortage during the day is 1 – F(S*) = 0.96.25 – 0.805) = 0.02 per day.

65. that only two spares should be brought. we find F(0) = 0. This occurs for S* = 2 which means. the purchase cost of each component. If more items are purchased to increase the stock to a level S. it may be less expensive to purchase no additional items than to order up to some order level. c = 10.323.000.135. the threshold is F(S*) = p –c 75–10 = = 0. a fixed ordering charge K is expended. Single Period Model with a Fixed Ordering Cost When the merchant has an initial source of product and there is a fixed cost for ordering new items. PO(z. we select the smallest value of S such that the CDF exceeds 0. then the assumption of the linear cost of lost sales would be violated.65. called the reorder point.857. p = 75.677. we assume that initially there are z items in stock.406. Such a policy is called the reorder point. Expressed in thousands. p + h 75 + 25 23 From the cumulative Poisson distribution using a mean of 2. We want to determine a level s. order level system. F(2) = 0. The expression for the expected profit is the same as developed previously. Because this is a discrete distribution. In this section. F(1) = 0. such that if z is greater than s we do not purchase additional items.000. If the system simply stopped after the spares were exhausted and a single cost of failure were expended. S) system. the extra cost of storage. except we must subtract the ordering charge and it is only necessary to purchase (S – z) items. or the (s. We consider first the case where additional product is ordered to bring the inventory to S at the start of the period.Single Period Stochastic Inventories h = 25.000 the cost of resupply. somewhat surprisingly. F(3) = 0. S) = b – c(S – z) + aEe[S] – (d + b)Es[S] – K (33) . The relevance of this model is due in part to the resupply aspect of the problem. This is in addition to the component initially installed. so that only on the third failure will a resupply be required. The probability of one or more resupply operations is 1 – F(2) = 0.

When z equals S. the profit when we replenish the inventory up to the level S is PO(z. we have plotted these the costs with and without an order. he should not restock. The two expressions are equal when z equals s. The higher of the two curves in Fig. When the demand has a normal distribution. If he has more than 210 papers. The two curves cross at about 210. PO(s*. As z decreases. The expected profit in this case is PN(z) =b + aEe[z] – (d + b)Es[z]. however. In Fig. This is the reorder point for the newsboy. Neither z nor K affect the optimal solution. but rather operate with the items on hand the profit is PN(z) = E[Profit] = b Here z = + kz . the optimal reorder point. The profit for a given day depends on how many papers the boy starts with. + a[z – + G(kz)] – (d + b)[ G(kz)]. he should order enough papers to bring his stock to 291.24 Inventory Theory We include the argument S with Ee[S] and Es[S] to indicate that these expected values are computed with the starting inventory level at S. S) = PN(s*) Generally it is difficult to evaluate the integrals that allow this equation to be solved. PN is greater than PO by the amount K. 8. z. (31) We modify the newsboy problem by assuming that the boy gets a free stock of papers each morning. . The profit is low when the initial stock is low and we do not reorder. Assuming a normal distribution and given the initial supply. (34) where the expected excess and shortage depend on z. As expected the profit grows with the number of free papers. If he has 210 papers or less. 8 shows the daily profit if one follows the optimal policy. If we choose not to replenish the inventory. and certainly no additional items should be purchased. and as before b–c+d F(S*) = b – a + d If no addition items are purchased. the system must suffice with the initial inventory z. PN and PO become closer. Then the optimal reorder point is s* where. the expected profit in the two cases can be written as a function of the distribution parameters. S) = b – c(S – z) + a[S – + G(k)] – (d + b)[ G(k)] – K (35) Here S = + k . The question is whether he should order more? The cost of placing an order is $10.

the lost income associated with a lost sale h = –$20. All remaining are disposed of at this price. If the inventory is not sufficient. a delivery fee will consist of a fixed charge of $500 plus $10 per item ordered.0 Reorder Not reorder 20. Determining the reorder point for the newsboy problem Example 6: Demand with a Uniform Distribution The demand for the next period is a random variable with a uniform distribution ranging from 50 to 250 units.0 Profit 30. such that . The current level of inventory is 100 units. sales are lost. the negative of the salvage value of the product. with a penalty equal to the selling price of the item. Items unsold at the end of the period go "on sale" for $20. however.Single Period Stochastic Inventories 25 70.0 50. c = $110. Additional items may be ordered at this time. (29). the order level is S.0 10. From Eq. and if so. the purchase cost plus the variable portion of the delivery fee K = $500. The selling price is $150.0 100 120 140 160 180 200 220 240 260 280 300 Initial Stock Figure 8.0 40.0 0. The purchase cost of an item is $100. Should an order be placed.0 60. the fixed portion of the delivery fee p = $150. how many items should be ordered? To analyze this problem first determine the parameters of the model.

1 ⌠ (S – A)2 Ee[S] = (B – A)⌡(S – x)dx = 2(B – A) A S 1 (B – S)2 ⌡ Es[S] = (B – A)⌠(x – S)dx = 2(B – A) S B CO = c(S – z) + K + h(S – A)2 + p(B – S)2 2(B – A) When no order is placed. CN = h(z – A)2 + p(B – z)2 .3 = 0 .3077.26 Inventory Theory F(S*) = p – c 150 – 1 1 0 = = 0. Expressing CN entirely in terms of z. 2(B – A) Evaluating CO with the order level equal to 112.05(s – 50)2 + 0.3077 or S = 111.729 – 110z. we solve for the optimal reorder point.375(250 – s)2 0. 19729 – 110s = –0. we find that CO = 19. we select S* = 112.05(z – 50)2 + 0.5s + 3543. S – 50 F(S) = 250 – 50 = 0. `Modifying the expected cost function to include the initial stock and the cost of placing and order. CO = c(S – z) + hEe[S] + pEs[S] + K For the uniform distribution ranging from A to B.325s2 – 72. Rounding up. p + h 1 5 0 –20 Setting the CDF for the uniform distribution equal to this value and solving for S. CN = –0.5.375(250 – z)2 Setting CO equal to CN. substituting s for z. the purchase cost and the reorder cost terms drop out and z replaces S.

Single Period Stochastic Inventories Solving the quadratic1 we find the solutions s = 150.3077.814. If there were no fixed charge for delivery. for s = 72. so we select the reorder point of 72. we get S = – [ln(1 – 0.3077)] = 55.17. The difference between s and S for the exponential distribution is approximately ∆=S–s= 2 K = c+h 2(150)(500) = 41 100 − 20 s = 56 – 41 = 15 For this distribution of demand.8 and s = 72. no additional inventory should be purchased. 1 The solution to the quadratic ax2 + b x + c = 0 is x = –b ± b2 –4a c] . the order would be for 12 units.3. 2a .3. At the optimal order level F(S*) = 1 – exp(–S/ ) = 0. the current inventory of 100 is considerably above both the reorder point and the order level. Certainly an order should not be placed. 27 Because the current inventory level of 100 falls above the reorder point. Example 7: Demand with an Exponential Distribution Consider the situation of Example 6 except that demand has an exponential distribution with a mean value = 150. we have CN = CO = 11. The solution lying above the order level is meaningless. Solving for S. At this point.

we consider the (s. Inventory Level Q r L L L L L 0 0 Time Figure 9. it is likely that only one order is outstanding at any one time. This is also called the stockout event. If we assume that L is relatively small compared to the expected time required to exhaust the quantity Q. The cycle begins with the receipt of the lot. alternatively called the reorder point. an order is placed for a lot size. L. This is called continuous review. The only uncertainty is associated with demand. We assume shortages are backordered and are satisfied when the next . The model assumes that the inventory level is observed at all times. but in reality the inventory decreases in a stepwise and uneven fashion due to the discrete and random nature of the demand process. s. Q) inventory policy.5 The (s. but allow the demand to be stochastic. that is the event of the inventory level falling below zero. Figure 9 shows the inventory pattern determined by the (s. we show the decrease in inventory level between replenishments as a straight line. we are most concerned with the possibility of shortage during an order cycle. order quantity system. Q. As we see in the figure. and then it continues for the time L when the next lot is received. There are a number of ways one might operate an inventory system with random demand. In Fig. In the following analysis. 9. it progresses as demand depletes the inventory to the level s. This is the case illustrated in the figure. The order arrives to replenish the inventory after a lead time. The lead time is assumed known and constant. When the level declines to some specified reorder point. Q) inventory policy. Q) Inventory Policy We now consider inventory systems similar to the deterministic models presented in Section 25.2. the inventory level increases instantaneously by the amount Q with the receipt of an order. We call the period between sequential order arrivals an order cycle. At this time. Inventory Operated with the reorder point-lot size Policy Model The values of s and Q are the two decisions required to implement the policy.28 Inventory Theory 25.

In practical instances the reorder point is significantly greater than the mean demand during the lead time so that Ps is quite small. With the average rate of demand equal to a. defined as Ps. one need only be concerned about the random variable that is the demand during the lead time interval. To determine probabilities of shortages. is Ps = P{x > s} = ∫ f (x)dx = 1 – F(s). It is the expected inventory level at the end of an order cycle (just before a replenishment arrives). Inventory Level Q s SS 0 0 L L L L Time Figure 10. The safety stock. s ∞ The service level is the probability that the inventory will not be depleted during one order cycle. This is the inventory maintained to protect the system against the variability of demand. The random demand during the lead time gives rise to the possibility that the inventory level will be depleted before the replenishment arrives. This is seen in Fig. and CDF F(x). This figure will also be useful for the cost analysis of the system. This probability.The (s. The (s. where we show the (s. Q) Inventory Policy 29 replenishment arrives. 10. is defined as SS = s – . f(x). The mean and standard deviation of the distribution are and respectively. Q) policy for deterministic demand . Q) policy for deterministic demand. SS. This is the random variable X with pdf. or Service level = 1 – Ps = F(s). the mean demand during the lead time is = aL A shortage will occur if the demand during the period L is greater that s.

and we observe that it is a function of the reorder point s. Q) = h( Q +s– ) 2 Ka + Q Inventory cost Replenishment cost . When we assume that the event of a stockout is rare and inventory declines in a continuous manner between replenishments. a practical assumption in many instances. The principal assumption is that stockouts are rare. The model is approximate in that we do not explicitly model all the effects of randomness. so the expected replenishment cost per unit time is Expected replenishment cost per unit time = Ka . the holding cost per unit time is Expected holding cost per unit time = h( With the backorder assumption. Q With the (s. The model and its optimal solution depends on the assumption we make regarding the cost effects of shortage. Q) policy and the assumption that L is relatively smaller than the time between orders. Q) policy. The cost for replenishment is K. EC(s. Q/a. the average inventory is approximately Average inventory level = Q +s– . Q) Policy We develop here a general cost model for the (s. We call this Cs. we use a as the time averaged demand rate per unit time. 2 Because the per unit holding cost is h. 2 Q + s – ). Combining these terms we have the general model for the expected cost of the (s.30 Inventory Theory General Solution for the (s. Dividing this cost by the length of a cycle we obtain a Expected Shortage cost per unit time = Q Cs. the shortage cost per cycle depends only on the reorder point. Since demand is a random variable.2. the time between orders is random with a mean value of Q/a. We investigate several alternatives for the definition of this shortage cost. In the model we use the same notation as for the deterministic models of Section 25. Q) policy.

there is a cost 1 expended whenever there is the event of a stockout. . (39) with Eq. Taking the partial derivative with respect to the variable s.   s  Now the partial derivative of Eq. We consider four different cases in the remainder of this section2. ∂s Combining Eq. Q) Inventory Policy a Cs Q 31 + Shortage cost (37) There are two variables in this cost function. with respect to each variable and set them equal to zero. we have for the optimal value of s ∂Cs hQ = – 1f(s*) = − . Eq. just on the fact that a stockout has occurred.   ∂s or ∂Cs hQ =– ∂s a (39) The solution for the optimal reorder point depends on the functional form of the cost of shortage. (37). Chapter 7. a ∂Cs ∂EC = h + Q  ∂s  = 0. Q and s. First. To find the optimal policy that minimizes cost.The (s. This cost is independent of the number of items short. we take the partial derivatives of the expected cost. The expected cost per cycle is ∞  Cs = 1P{x > s} = 1  ∫ f (x)dx . the partial derivative with respect to Q is ∂EC h a(K + Cs ) = – =0 ∂Q 2 Q2 or Q* = 2a(K + Cs ) h (38) We have a general expression for the optimal lot size that depends on the cost due to shortages. ∂s a 2In (40) this article we follow the development in Peterson and Silver [1979]. Case of a Fixed Cost per Stockout In this case. (40). (40) with respect to s is ∂Cs = – 1f(s).

We adopt a continuous review policy in which the order quantity is the average demand for one month.32 Inventory Theory hQ . It can be shown that the cost function is minimized when f(x) is decreasing. We must first adopt a time dimension for those data items related to time. Find the optimal inventory policy.01) = $10/unit-month. The pdf may satisfy this condition at two different values. Holding cost is estimated using the interest cost of the money invested in a unit of inventory. a = 100 units/month h = 1000(0. 1a or f(s*) = (41) (42) and Cs = 1(1 – F(s*)). Equation (41) is a condition on the value of the pdf at the optimal reorder point. the order cost. Here we use 1 month. the cost of paperwork is estimated to be $200. If both parameters are flexible. is used to find values of Q and s that solve the problem. The fixed order cost is $800. for Q* define Example 8: Optimal reorder point given the order quantity ( 1 Given) The monthly demand for a product has a normal distribution with a mean of 100 and a standard deviation of 20. If no values of the pdf satisfy this equality. so for a unimodal pdf. We must also describe the distribution of demand during the lead time. select some minimum safety level as prescribed by management. (38) the optimal control parameters. For convenience we assume that 1 month has 4 weeks and that the demands in the weeks are independent and identically distributed normal variates. The lead time for this situation is 1 week. The purchase cost of the product is $1000. Equation (41) specifying the optimal s* together with the Eq. the unit cost multiplied by the interest rate (interest rate is 12%/12 = 1% per month) 1 = $1000. For this selection. The interest rate used for time value of money calculations is 12% per year. which is independent in time and number K = $800. as illustrated in Example 13. select the greater of the two solutions. When it is necessary to backorder. If one of the parameters are given at a perhaps not optimal value. the backorder cost. a successive approximation method. With these assumptions the weekly demand has . independent of the number backordered. these equations yield the optimum for the other parameter.

the expected shortage is Es = ∫ (x – s) f (x)dx .The (s. Taking the larger of the two possibilities we find s* = + (1. thus Q = 100. 1a The pdf of the standard normal distribution is related to a general normal distribution as f(s) = (1/ ) (k) or (k) = f(s) Then in terms of the standard normal we have (k*) = hQ = (10)(0.6 or 42 (conservatively rounded up). we may also be interested in the expected number of items backordered during an order cycle. Q) Inventory Policy = 100/4 = 25. 0. if x ≤ s Items backordered = x – s.01.   ∂s s  .1. 1a We look this up in the standard normal table provided at the end of this chapter to discover k* = ±1. This depends on the demand during the lead time.66) = 25 + 1. Es. (41). we find the associated optimal reorder point. if x > s  Therefore. and 2 33 = 10.66. or f(s*) = (10)(100) hQ = (1000)(100) = 0. Using this value in Eq.01) = 0. This is the optimal reorder point for the given value of Q. we assume a cost 2 is expended for every unit short in a stockout event. s ∞ For this situation. The expected cost per cycle is Cs = 2Es. = 202/4 = 100 or The problem specifies the value of Q as 1 month's demand.66 (10) = 41. Case of a Charge per Unit Short In some cases. Now the partial derivative with respect to s is ∞  ∂Cs = – 2  ∫ f (x)dx0  = – 2(1 – F(s)).

95. The service level is actually the value of F(s).34 Inventory Theory From Eq. but change the cost structure for backorders. (25).  . From the normal distribution table we find that this is associated with a standard normal variate of z = 1. The optimal policy is governed by Eq. ∞  Cs = 2Es = 2  ∫ (x – s*) f (x)dx  . a 2 or (43) In this case. Now we assume that we must treat each backordered customer separately.64. (38) by substituting the value of Cs. One might require that the inventory meet demands from stock in 99% of the inventory cycles. The cost of paperwork and good will is estimated to be $200 per unit backordered. the optimal order quantity is determined from Eq. Managers may find it difficult to specify the shortage cost 2. (41).   s  (44) This integral is difficult to compute except for simple distributions. The reorder point is then  . Example 9: Optimal reorder point given the order quantity ( 2 Given) We consider again Example 8. If the expression on the right is less than zero. use some minimum reorder point specified by management. It is easier to specify that the inventory meet some service level. Given values of h. F(s*) = 1 – (10)(100) hQ = 1 – (200)(100) = 0. (43) the implied shortage cost for the given service level. Q and a.95. 2a We know that the probabilities for a normal distribution is related to the standard normal distribution by s– F(s) =   (k*) = 0. This is 2. (43). one can compute with Eq. It is evaluated with tables for the normal random variable using Eq. For a given value of s. the optimal value of s must satisfy ∂Cs hQ = – 2(1 – F(s*)) = − ∂s a hQ F(s*) = 1 – . we have a condition on the CDF at the optimal reorder point.

The expected cost per cycle is Cs = 3Ts.64 (10) = 41. the average time a customer must wait for delivery is x–s .The (s. This is the optimal for the given value of Q.4 35 or 42 (conservatively rounded up). Case of a Charge per Unit Short per Unit Time When the backorder cost depends not only on the number of backorders but the time a backorder must wait for delivery. 2a ∞ (45) We consider here the case when a cost 3 is expended for every unit short per unit of time. 2a The resulting unit-time measure for backorders is (x – s)2 . [1998] Qh report the result Es(s*) = which is derived using a more accurate h+ 3 . (41).   ∂s a s a  From Eq. we would like to compute the expected unit-time of backorders for an inventory cycle. Q) Inventory Policy s* = + (1. 2a Integrating we find the expected value Ts. (47) We have added the parameter s* to the expected shortage to indicate its value is a function of the reorder point. the optimal value of s must satisfy ∂Cs E hQ = − 3 s =– ∂s a a or Es(s*) = hQ 3 (46) . When the number of backorders is x – s and the average demand rate is a. Note that Silver et al. where 1 2 Ts = ∫s (x – s) f (x)dx . Now the partial derivative of Cs with respect to s is ∞  ∂Cs E = – 3  ∫ (x – s) f (x)dx  = – 3 s .64) = 25 + 1.

When stock is exhausted during the lead time. This is 3.9) = 25 + 9 = 34 This is the optimum for the given value of Q. The two results are approximately the same when 3 >> h. Example 10: Optimal reorder point given the order quantity ( 3 Given) We consider again Example 8. the expected shortage at the optimum is Es(s*) = G(k*) = 1 where k* = s* − or G(k*) = 0. Es. We use L to indicate the cost for each lost sale. (47). The reorder point is then s* = + (0. Es(s*) = hQ 3 = (10)(100) = 1. Lost Sales Case In this case sales are not backordered.9. The effect of this situation is to raise the average inventory level by the expected number of shortages in a cycle. 1000 When the demand is governed by the normal distribution. For the case of lost sales the approximate expected cost is Q EC(Q. as assumed here.1 From the table at the end of the chapter k* = 0. but now we assume that $1000 is expended per unit backorder per month. The optimal policy is governed by Eq. s) = h  + s − 2 + aK Q  + Es   Inventory cost Replenishment cost . A customer that arrives when there is no inventory on hand leaves without satisfaction. and the sale is lost. the inventory level rises to the level Q when it is finally replenished.36 Inventory Theory representation of the average inventory. We also experience a shortage cost based on the number of shortages in a stockout event.

This is L. These are summarized below for easy use. Q) Inventory Policy a 37 + L Q Es Shortage cost (48) Here we are neglecting the fact that with lost sales. We charge $2000 for every lost sale. but now we assume that the sale is lost given a stockout.995 hQ + L a L a From the table at the end of the chapter k* = 2. The number of orders per unit time is slightly less than a/Q. not all the demand is met. The optimal policy is governed by Eq.The (s. Taking partial derivatives with respect to Q and s we find the optimal lot size is Q* = 2a(K + h L Es )  ∂E s    = 0. The reorder point is then s* = + (2.6 = 47.58) = 25 + 22.6 This is the optimum for the given value of Q. Summary We have found in this section solutions for several assumptions regarding the costs due to shortages. F(s*) = (2000)(100) = (10)(100) + (2000)(100) = 0.58. (50). The optimal reorder point requires one to find the value s* that corresponds to .  ∂s  (49)  ∂EC ∂E  a L = h  1 + s  )+  ∂s ∂s  Q or ∂Es hQ =− ∂s hQ + L a (1 – F(s*)) = hQ hQ + L a (50) or F(s*) = 1 – hQ La = hQ + L a hQ + L a Given) Example 11: Optimal reorder point given the order quantity ( L We consider Example 8 again.

and is given by Q* = 2a(K + Cs ) h This equation is used directly when a value of s is specified. F(s*) or Es(s*) equaling some simple function of the problem parameters. what is the optimal order quantity? For a normal distribution the expected shortage cost is Cs = 2 G(ks ) . Cs. Q) Policy for Continuous Distributions Situation Fixed cost per stockout ( 1) Charge per unit Short ( 2) Charge per unit short per unit time ( 3) Charge per unit of lost sales ( L) 1(1 Cs – F(s)) 2Es 3Ts Optimal reorder point f(s*) = hQ 1a hQ 2a Normal solution (k*) = (k*) = 1 – G(k*) = hQ 1a hQ 2a F(s*) = 1 – Es(s*) = hQ 3 hQ 3 LEs F(s*) = L a L hQ + a (k*) = L a L hQ + a Determination of the Order Quantity Up until this point. The demand during the lead time is normal with µ = 25 and = 10. The following examples illustrate the determination of the order quantity when the reorder point is given. If the reorder point is fixed at 50. Example 12: Optimal order quantity given the reorder point We continue from Example 9 in which the shortage cost is 2 = $200 per unit short. Table 1. The (s. The optimal order quantity for each case depends on the shortage cost. and the determination of optimal values for both variables simultaneously.38 Inventory Theory f(s*). It is used iteratively when the optimum for both s and Q is required. all our examples have determined the reorder point given the order quantity.

5.53) = (200)(10)(0. We need an iterative procedure to find both. Q= 2a(K + Cs ) . Q) Inventory Policy where ks = (s – )/ .5.8 10 39 or 127 (conservatively rounded up). G(2.53 or s = 40. we have Q2 = 130.5) = 0. Using this value of Q2. h (ks) = 1 – hQ . Cs = 4. Example 13: Both optimal order quantity and reorder point In the previous examples we fixed one of the decisions and found the optimal value of the other. . For s = 50 and ks = 2.1 Computing the associated Cs we find Q3 = 130.7.72 For this value of Cs.The (s.02736) = 54. so we adopt the policy Q* = 131 and s* = 40. Es = 0. Q1 = 126. Q* and s*. It appears that the values are converging. we find the optimal reorder point ks = 1.51 or s = 40. Cs = 2 G(ks ) a 2 The first step is to assume Cs = 0 and to find the corresponding optimal order quantity. we find the optimal reorder point ks = 1. Then the optimal order quantity is Q* = 2a(K + Cs ) = h 2(100)(800 + 4) = 126. We use the expression below sequentially. Using this value of Q1.9.020.0020.3 The expected shortage per period with this reorder point is Cs = 2 G(1.

11. A move practical idea is to use the inventory position rather than the inventory level as an indicator. The inventory position is the inventory level plus the quantity on order. Using inventory position as a measure for placing orders Using the inventory position in this manner. and the same order pattern is obtained using either measure. This probability depends on the reorder point and not the lot size. however. In this case the inventory level falls below the reorder point and never reaches it again. The difference is illustrated in Fig. Basing the order on the inventory level fails. however. as in the last cycle of the figure. Using the inventory position. also allows us to drop the requirement that the lot size be very much greater than the average demand during the lead time. The primary assumption for the derivations is that the probability of a stockout be small. the inventory level crosses the reorder point at the same time as the inventory position. thus keeping the inventory in control. . allows two orders to be placed in quick succession.40 Inventory Theory 25. when there is a lead time demand larger than the order quantity. We note that inventory level is the same as inventory position when there are no outstanding orders.6 Variations on the (s. Q) Model Reorder Point Based on Inventory Position In the foregoing. The results in the table can be used even in cases where the lot size is small in relation to the lead time demand. we have assumed that a replenishment order is to be placed whenever the inventory level reaches the reorder point. In the early cycles of the figure. Q s L L L L L 0 0 Time Inventory Position Inventory Level Figure 11.

Variations on the (s. The expected shortage and expected unit-time shortage are Es = x =s +1 ∑ (x – s) p(x) ∞ (51) Ts = 1 ∞ ∑+1(x – s)2 p(x) 2a x =s (52) Table 2. Q) Model 41 When the lot size is small. Indeed this is a very desirable characteristic of an inventory system. This is particularly true when the reorder point is relatively small. For the discrete distribution p(x) is the probability that the random demand during the lead time takes the value x. Discrete Demand During the Lead Time The results of the table were derived for continuous distributions. there may be many outstanding orders at any given time. This implies that a replenishment order is placed whenever an item is withdrawn from inventory. emphasizing the need to track the inventory position. Q) Policy for Discrete Distributions Situation Fixed cost per stockout ( 1) Charge per unit short ( 2) Charge per unit short per unit time ( 3) Charge per unit of lost sales ( L) 1(1 Cs – F(s)) 2Es 3Ts Optimal reorder point p(s* – 1) > F(s*) ≤ 1 – Es(s*–1) > hQ ≥ p(s*) 1a hQ < F(s*+1) 2a hQ 3 ≥ Es(s*) LEs F(s*) ≤ 1 – hQ < F(s*+1) hQ + L a Lead Time a Random Variable Previously we assumed that lead time is a constant. A particularly interesting case is when the lot size is 1. F(x) is the probability that the demand is less than or equal to x. The lead time may actually be uncertain in duration due to variability in shipping times. The (s. In fact the items in an inventory are usually discrete. material availability and supplier processing times. and a discrete demand distribution may be more appropriate. .

The marginal distribution of demand during the lead time is f(x) = ∫ g(x.42 Inventory Theory Let lead time be a random variable Y with pdf h(y).that is. . y)h(y)dy 0 ∞ (53) This pdf can then be used in conjunction with the standard normal distribution table to determine approximate solutions. and let demand be the random variable X with joint pdf g(x. the demand distribution depends on the lead time. y) -.

After a lead time interval L. the lowest inventory that is affected by that order occurs at time t + R + L. S) = h ( aR +S– ) 2 Inventory cost . For the (R. If the inventory is at level y. S) Inventory Policy A different way to manage a stochastic inventory system is illustrated in Fig. S) policy is much more affected by variability than the (s. For the (s. we define the demand in the interval R + L to be the random variable X. Q) policy. The mean and variance during the interval are P and P (the subscript “P” stands for periodic). the reorder point s is set to protect against the possibility of shortage during the lead time L. In the event of a particular order at time t. S) policy. S) Inventory Policy 43 25. The (R. the replenishment order is delivered. Q) policy. 12 This is called a periodic review policy in that the inventory level is only observed at intervals of R.7 The (R. S) inventory policy The analysis of this policy is much like that for the (s. S is called the order level. The (R. Q) policy because of the longer interval. respectively. The quantity S must be large enough to keep the probability of a shortage in that time interval small. a quantity S – y is ordered to bring the inventory position to S.The (R. The cost (per unit time) of operation of the inventory system expressed in terms of R and S is EC(R. To analyze this system. Figure 12 shows the inventory position with dotted lines and the inventory level with solid lines. the order level S is set to protect against a shortage in the time interval R + L. S 0 0 L R L R L R L Time Figure 12. The advantage of the policy is that it does not require continuous review. The pdf and CDF are fP(x) and FP(x).

the determination of the optimal order level follows exactly the derivation for s* for the (s. The optimal conditions are found in Table 1. . We create Table 3 from Table 1 by replacing s with S. and L with R + L. Q with aR. Q) policy. In particular the distribution of demand in the expressions for Cs have a mean and variance that depends on the interval L + R. If we assume that the review interval R is set elsewhere. A similar table for discrete distributions can be constructed starting from Table 2.44 Inventory Theory K R 1 + CS R + Replenishment cost Shortage cost (54) Evaluation of the shortage cost depends on the assumption of the costs experienced in the event of a stockout. ∞  Fixed cost per stockout: Cs = 1(1 – F(S)) = 1  ∫ fP (x)dx   S  Charge per unit short: ∞  Cs = 2Es = 2  ∫ (x – s) fP (x)dx   S   1 ∞  Cs = 3Ts =  ∫ (x – s)2 fP (x)dx  2a    S Charge per lost sale: Cs = LEs (55) (56) Charge per unit short per unit time: (57) = L ∞   (x – s) fP (x)dx ∫  S  (58) We cannot determine the optimal R easily in this case because Cs depends on R.

the cost of paperwork and good will is estimated to be $200 per unit. The monthly demand for the product has a normal distribution with a mean of 100 and a standard deviation of 20. When it is necessary to backorder a customer request. S) Inventory Policy Table 3. the cost of holding one unit for one month 2 = 200. the cost of backordering one unit = 20 during the review period (the distribution is = 100 and normal) With these values we can determine the optimal policy. Find the optimal inventory policy.957 for z = 1. we find that the CDF has the value 0. (k*) = 1 – hR 2 = 1 – (10/200) = 0.The (R.65. The (R. The order level is S = 100 + 20(1. S) Policy for Continuous Distributions Situation Fixed cost per stockout ( 1) Charge per unit Short ( 2) Charge per unit short per snit time ( 3) Charge per unit of lost sales ( L) Example 14 1(1 45 Cs – F(S)) 2Es 3Ts Optimal reorder point fP(S*) = hR 1 Normal solution (k*) = P hR 1 FP(S*) = 1 – Es(S*) = hR 2 (k*) = 1 – G(k*) = hR 2 haR 3 haR P 3 LEs F(S*) = 1 – hR hR + (k*) = 1 – L hR hR + L We consider again the situation of Example 9 except the inventory is reviewed every month. The parameters of the model are as follows R = 1 month h = $10. The holding cost is $10 per unit per month.65) = 133. The lead time for orders is zero. .95 Using the standard normal table.

we find that the CDF has the value 0. . The probability of a shortage during the month is 1 – (k*) = 0. Three extra units of inventory are necessary for the longer review period.28.05.90.28) = 236. The safety stock is S – µ = 33 units.46 Inventory Theory The optimal policy is to order a quantity (133 – x) each month.28 during the review period (here we assume that the monthly demands are independent) (k*) = 1 – hR 2 = 1 – (10)(2)/200) = 0. Now we consider the same problem when the review period is every 2 months. The situation is the same except R = 2 months = 200 and = 20 2 = 28. The safety stock is: S – µ = 36. The optimal policy is to order a quantity (236 – x) every two months. Thus the reorder level is S = 200 + 28.899 for k = 1.28(1. For the continuous review of Example 9 the safety stock is 17 units. Using a standard normal table.

(b). what is the optimal policy? e.15)(5) = $0. It is now January 1. the inventory is replaced by a production process that can add units to the inventory at a constant rate r (r > a ). When should the next order be placed for the solutions of parts (b). A hardware distributor is reviewing his inventory policy for hammers. which have a relatively constant demand of 2000 units per month. f. An OR analyst suggests that money might be saved if the customers would accept backorders. Derive a formula for the optimal lot size when no shortages are allowed. so the cycle time must be in whole months. Each time the process is begun a setup cost K is expended. Every time he places an order for replenishment a shipping and paper preparation cost of $500 is charged. and (e) of this problem? g.2. What is the optimal lot size when no shortages are allowed? c. There is a lead time (time between when the order is placed and when it is delivered) of 1 month. The distributor's holding cost is 15% (annual) of his average investment in inventory.10 discount on the price of hammers for an order of at least 10. (c). offers a $0. (d). What is the optimal lot size in this case? d. Rather than an instantaneous replenishment of inventory as assumed in Section 25. The current inventory policy is to replenish the inventory every month. A meat market advertises that its hamburger is the freshest in town.75 per hammer per year. a. Should the distributor take advantage of this deal? Assume no shortages are allowed. The distributor buys the hammers from his supplier for $5 each and sells them for $10. (The holding cost is h = (0.Exercises 47 25. and (e). It is now January 1. What is the total annual cost of this policy? b. and the current inventory is 3500 units. The supplier. (d). Otherwise the inventory problem is the same as previously stated. The distributor wants to order only at the beginning of the month. If the distributor offers a discount of 1% of the purchase price for every day the customer has to wait for delivery. (c). To get rid of old stock.000 units.2 1. the price of the hamburger is reduced by $0. Of course. the process cannot be operated continually but must be started and stopped.) Answer each of the following independently ( one part does not depend on another). in an effort to get larger orders. . No shortages are al- 3.10 per pound for every hour it remains unsold after it is ground (for simplicity assume this reduction is linear in time).8 Exercises Section 25. Compute the annual profit on hammers for parts (a). The demand for hamburger is 250 pounds per hour. 2.

She can order a case that holds 12 units for a cost of $1200 or a gross that holds 144 units for $13.000. the sales are lost. A principal concern for this business woman is the cost of capital. how often will bulbs be delivered? c. so the alternatives represent ordering a 1-month or 1-year supply. but the actual demand has a Poisson distribution. regardless of the number of cases delivered. The expected demand for trees is 3 per day. The demand for the product is 12 units per month. what penalty is implicitly being charged for a shortage? For this value find 8. a. Management has determined that this is too high. the purchase cost is not relevant. The management of the building uses a 15% annual carrying cost rate for inventory. Trees remaining at the end of the season must be cared for until the next year. A case of bulbs costs $200. A long-standing policy has been to receive 10 cases per delivery. A garden nursery has a short selling season of 20 days for a certain kind of tree. Only the shortage cost and holding cost should be used. An office building uses bulbs at an average rate of 1000 bulbs per month. Section 25. If all other parameters remain the same. These costs include delivery and product cost. (Note that since leftover trees will actually be kept after the season and eventually sold.48 Inventory Theory lowed. Which plan should she adopt? Light bulbs come in cases of 144 bulbs. The company that sells the bulbs charges $50 per delivery.) For the parameters given in Example 6 in the text. The management assumes a "loss of good will" charge of $10 per bulb per month for backorders. the probability of a shortage is almost 70%. What is the annual cost penalty of this nonoptimal policy over the cost of the optimal policy? 5. As an approximation assume that demand is at a constant rate and no shortages are allowed. which is 20% per year. Find the number of trees that should be ordered and the probability that this inventory level will be sufficient to meet all demand. Consider a gift shop operator with a relatively constant demand for a certain type of wall hanging. independent of the amount processed. A level of 10% is considered an acceptable probability of shortage.4 7. The cost of such care per tree is $20 and the holding cost on investment is $2. No shortages are allowed. Find the optimal policy for 5 if backorders are allowed. 6. Assuming the usage rate for bulbs is constant. With this order quantity. b. The nursery currently has 20 such trees in stock and has one last opportunity to order more. how many cases of bulbs should be ordered in each delivery? Only whole cases can be ordered. If demand occurs after the inventory is exhausted. A failed bulb is simply not replaced until the beginning of the next month if the supply is exhausted. How much should be ground at each setup? 4. The cost to set up and run the hamburger grinder is $100. . The nursery purchases the trees for $75 and sells them for $175.

he charges a lost opportunity cost of $10.02 9 0. How many units should he build to maximize his expected profit? 11. In the table below. p = $100.000 each. while it will cost only $40. Use the penalty found in Exercise 8. How many helicopters should the commander bring? 12.02 7 0. x is the number of units demanded and p (x) is the probability of that demand. A military commander faces a dangerous mission which requires the use of helicopters. The mission lasts 10 hours.14 12 0.000 to build it. Consider a problem with the following parameters.09 11 0. h = $1. Find the optimal (s . Failure-causing events are assumed to occur at random on the average of once every hour.000.03 6 0.05 3 0.10 15 0.14 14 0. he will have to dump them for $20.000. The helicopters cost $1 million each. Demand has a uniform distribution with A = 20 and B = 60 b. The number of events is independent of the number of helicopters active. His problem is to determine how many units to build.03 10 0. and K = $200.05 The selling price for a condo unit is $100. Demand has an exponential distribution with a mean value of 40 . after which he will leave town for places unknown.04 4 0.19 13 0. and find the optimal policy if the demand is exponential distributed with a mean of 150. x p (x) x p (x) 1 0. the commander has attached a cost of $100 million for every unit less than five that finishes the mission. In order to quantify the situation. Each event will destroy one helicopter.01 8 0. If the developer runs out of units while there is still demand. The commander wants to determine how many to bring on the mission.04 5 0. S) policy when: a. Any units that remain at the end of the mission are destroyed. 49 10. He estimates the following probabilities for the demand for his units. 9.Exercises the optimal policy assuming the distribution of demand is uniform between 50 and 250 items. A dishonest developer wants to make quick money building a condominium. At least five are required for mission success. He sets a 1-year time table for his activities.05 2 0. If the year passes and he has not sold all his units. Any number less than five is judged a serious detriment to the mission. c = $10.

2 = $200. h = $120/unit-year. The average demand doubles but the standard deviation remains the same. 16. The order quantity is set to the average demand for 1 month. The backorder cost is $5 per unit. a = 1200 units/year. Find the optimal continuous review policy for the following situations. Q) policy when the following changes are made. The fixed setup cost doubles. What should be the reorder point if the lead time were 1 month? . Use the data of Exercise 14 and add the information that the setup cost is $200 and the lead time is 1 week. The holding cost doubles. Find the optimal order level for an inventory system for which orders are placed every 2 weeks. What should be the supplier's reorder point if she wants a 60% chance of not having a shortage during the lead time? b.7 Inventory Theory 13. b. The standard deviation of demand per week doubles. The changes are not cumulative. d. The cost of the item is $20.5 . Customer arrivals follow a Poisson process.25. a. The engines are obtained from a manufacturer who delivers 2 months after an order is placed. The backorder cost doubles. The demand per month has a normal distribution with mean 100 and standard deviation 20. K = $800. b. Assume four weeks per month. a. A supplier of rebuilt engines expects an average of three engine customers every 2 months. c. 15. a. e. The lead time doubles. Interest per year is 20%. 14. f. c. Find the optimal (s. The reorder point is set at the average demand over the lead time. Neither the reorder point nor the order quantity is specified. The lead time is one week.50 Section 25. We repeat the conditions of Example 9. The weekly demand has a normal distribution with a mean of 25 and a standard deviation 5. Assume a 50-week year.

Customers arrive at random (a Poisson process) so the exact demand for any given period of time cannot be computed. a.5% per week. the standard deviation is 3000 gallons.75 per gallon. He must pay $0. The average demand is 10. What reorder point will provide a 95% chance that the inventory will not run out during an order cycle? b. except that demand is stochastic rather than deterministic. The weekly demand for light bulbs is a random variable with a normal distribution that has a mean of 250 bulbs and a standard deviation of 50 bulbs.000 gallons per week. but if any remains in inventory when the next order arrives he charges a holding cost based on this interest rate and the value of the inventory. The size of the order equals the previous month's sales. What kind of inventory system is this? b. Assume for simplicity that months are 4 weeks long and that there are 48 weeks in a year. paying an extra $0. He adopts a (s. A lot size of 100 is selected. If the supply of bulbs runs out and a bulb fails. the tenant simply must wait until the beginning of the next month. For purposes of analysis. His cost of capital is 0. What specific numerical rule should the management use to determine how much to order each month? 18. The manager in Exercise 18 installs computerized inventory control so that continuous review is possible. He recovers this cost in the price he charges for gasoline. An order for bulbs is placed once a month (on the first of the month) and the order is delivered right away. What is the optimal inventory policy for the distributor? 21.10 per gallon for the privilege. If a design change is to be incorporated into the . One of the disadvantages of using large order quantities in a production process is. If the distributor runs out of supply during the period. Q) policy. Another disadvantage is the inflexibility associated with having a large inventory. The borrowed gas must be returned when his supply is replenished. of course. large holding costs. however. His supply is replenished every 6 weeks. this event is assumed to have a cost of $10. Consider the light bulb situation of Exercise 5. What is the average cycle time for this plan? 20. he "borrows" gas from other distributors. The average demand for a product at a warehouse is 1200 units per year.Exercises 51 17. A gasoline distributor has a weekly demand that is approximately normally distributed. The warehouse manager replenishes the inventory by a monthly order to the factory. a. What order level will provide a 95% chance that the inventory will not run out during an order cycle? 19. Assume a month is 4 weeks. A lead time of 1 week passes between when an order is placed and when the inventory is replenished.

B = $10. Compute the optimal values of s and Q for each case and compare the total costs (including B) and production lead times. c. This is one of the tenants of the "just-in-time" production systems.000 K = $1. B = $110. Use a 50-week year and an (s. and standard deviation 10. This might be called the lead time of the production process. so reducing setup cost results in reduced holding cost. and different annual costs to implement the production process B. Q) inventory policy. The reason for a large order quantity is a large setup cost.000 K = $5. which go to great measures to reduce setup costs.000 . and increased flexibility. so as Q grows so does the length of the cycle. b.000 In each case weekly demand follows a normal distribution with mean 50.52 Inventory Theory product or some custom feature is to be added for particular customers.000 . K = $10. Choose the least cost system. The average length of the cycle is Q /a. a. the modification must await the next production cycle while the entire inventory is depleted.000 . B = $30. The lead time when a reorder is placed is 1 week. . The holding cost is h = $200/unit-year. decreased production lead time. Analyze the following production systems that are characterized by different setup costs K. The shortage cost is $100 per unit.

0596 0.0044 0.1394 0.70 -0.9202 0.05 -1.6978 0.0022 0.0283 0.60 -0.45 -2.7687 0.0122 0.3264 0.7011 2.1686 1.0668 3.5293 -1.0175 0.4828 1.3867 0.55 -2.90 -2.4027 2.0030 0.3085 0.1469 0.30 -2.65 -0.00 0.9597 1.55 -1.1556 2.0040 0.7183 1. CDF G(y) = (y) – y[1 – (y)] .3446 0.0154 0.3989 0.55 -0.8054 0.0416 1.4207 0.2542 2.95 -2.5000 1.3984 0.6512 2.0355 0.3970 0.0104 0.0940 0.6232 1.3521 0.0224 0.1109 0.0574 2. Function Values for the Standard Normal Distribution.90 -0.80 -2.0179 0.2661 0.15 -2.0026 0.65 -1.1604 0.0540 0.3006 1.5761 1.0062 0.0079 0.0119 0.5668 0.0446 0.3910 0.8008 2.35 -0.3821 0.80 -0.1497 0.9005 2.1942 0.0863 0.0202 0.8626 1.6637 0.25 -1.0606 0.3683 0.15 -0.0885 0.3011 0.0071 0.0016 0.20 -2.0035 0.1714 0.0317 0.10 -0.0790 0.0107 0.75 -0.3989 † y is a standard normal variate (y) is the probability density function.1257 1.3605 0.0656 0.45 -1.60 -2.1587 0.65 -2.6304 0.4367 1.3532 2.1295 0.2119 0.1151 0.85 -0.00 -1.0401 0.9600 0.10 -1.4013 0.0054 0.4509 0.0069 0.4244 0.0287 0.0256 0.2059 0.50 -0.0091 0.2179 0.45 -0.8429 0.2541 0.8506 2.9505 2.2299 0.1200 0.5020 2.0548 0.10 -2.50 -1.6015 2.4523 2.20 -0.0440 0.3332 0.2561 1.2121 1.75 -1.3945 0.7662 1.90 -1.7509 2.4801 0.05 -2.0094 0.0198 0.3230 0.0013 0.1056 0.5981 0. y ∈ [−3.0252 0.0968 0.0721 0.2912 0.8812 0.0004 2.2780 0.4602 0.50 -2.3814 0.0488 0.0495 0.95 -1.3037 2.4784 0.1251 0.5293 1.0735 0.1841 0.8143 1.30 -1.0158 0.6706 1.0139 0.75 -2.1826 0.9111 1.0060 0.2420 0.0808 0.25 -2.35 -2.3752 0.5363 0.5517 2.2420 0.0359 0.15 -1.70 -1.1977 0.3455 1.1065 2.85 -2.0004 0.1711 0.1357 0.3632 0.0833 1.60 -1.40 -2.95 -0. pdf (y) is the cumulative distribution function.Exercises Table 4.0047 0.0228 0.2743 0.35 -1.0051 0.3909 1.4404 0.2897 0.0136 0.20 -1.7328 0.2578 0.50 0.40 -1.3123 0.40 -0.5069 0.3429 0.80 -1.2266 0.2049 2.0] † y (y) (y) G(y) y (y) (y) G(y) 53 -3.0668 0.30 -0.70 -2.0085 1.1295 0.85 -1.0019 0.0322 0.00 -2.1023 0.05 0.0396 0.0082 0.00 -0.25 -0.

0005 0.90 0.8289 0.0440 0.50 1.9987 0.8023 0.8749 0.45 0.9713 0.1100 0.6736 0.5000 0.0111 0.0757 0.0162 0.1978 0.05 0.30 0.0023 0.0206 0.8159 0.0283 0.1200 0.9842 0.1394 0.00 1.0621 0.30 2.0183 0.3867 0.50 2.3989 0.0074 0.95 3.0175 0.10 1.9332 0.5987 0.0252 0.0037 0.0506 0.0328 0.70 1.0060 0.15 0.0833 0.2863 0.0317 0.8849 0.45 1.30 1.9505 0.85 2.70 2.7422 0.3910 0.0056 0.65 1.9641 0.0455 0.0091 0.9893 0.7257 0.8944 0.0017 0.3509 0.1004 0.6915 0.9878 0.75 2.5199 0.9452 0.35 2.05 2.1942 0.6554 0.0042 0.1828 0.10 0.0367 0.0009 0.2541 0.6179 0.0916 0.3123 0.3605 0.1429 0.0104 0.0008 0.60 0.40 2.7734 0.8643 0.3011 0.54 Inventory Theory Table 4.0656 0.40 1.0012 0.00 2.9265 0.0049 0.0561 0.0079 0.50 0.3332 0.9744 0.55 0.0409 0.3683 0.0085 0.9032 0.65 2.50 0.0224 0. (Cont.0143 0.3284 0.40 0.0863 0.0293 0.25 1.9938 0.3] † y (y) (y) G(y) y (y) (y) G(y) 0.3521 0.80 1.1554 0.0136 0.55 2.20 1.0721 0.70 0.10 2.0065 0.90 2.3989 0.0198 0.1023 0.9970 0.2299 0.3069 0.9974 0.15 2.0293 1.9332 0.9965 0.1295 0.1497 0.0488 0.1295 0.80 2.0396 0.7580 0.9984 0.00 0.9918 0.1312 0.0069 0.9861 0.0355 0.0097 0.0540 0.2661 0.2420 0.85 1.85 0.05 1.3945 0.0051 0.35 1.0686 0.20 0.75 0.5596 0.0044 0.60 2.2059 0.2897 0.0027 0.9192 0.9929 0.1687 0.75 1.0020 0.3752 0.9960 0.2780 0.9821 0.25 2.0232 0.1109 0.9981 0.9978 0.7088 0.3970 0.0006 0.2668 0.3814 0.00 0.9798 0.3429 0.55 1.25 0.3984 0.2304 0.0119 0.0790 0.0011 0.0004 .9772 0.2179 0.9953 0.5793 0.) Function Values for the Standard Normal Distribution y ∈ [0.15 1.95 2.80 0.1714 0.2481 0.90 1.1826 0.9906 0.0261 0.7881 0.3230 0.20 2.1604 0.6368 0.35 0.45 2.95 1.9554 0.0032 0.8531 0.3744 0.0126 0.65 0.1202 0.9678 0.8413 0.9115 0.0154 0.0015 0.60 1.0005 0.9946 0.9394 0.0940 0.0596 0.5398 0.2137 0.9599 0.

Foundations of Operations Management. Operations Management. . Englewood Cliffs.R. McGraw Hill. New York. Russell. 2000. Peterson.J. 2003. Peterson. L. Upper Saddle River. NJ. Englewood Cliffs. Boucher. Modeling and Analysis of Manufacturing Systems..J. NJ. Control and Integration.C. and M.G. R. R. 1994. D. John Wiley & Sons. Principles of Inventory and Materials Management. McGraw Hill.O. Chicago. Factory Physics. New York.A. 1979. Hopp. Upper Saddle River. and B. and D. Sipper. S. Prentice Hall. New York. Silver. New York. H. R. New York. John Wiley & Sons. John Wiley & Sons. Johnson. 1979.H. Analysis and Control of Production Systems. Ritzman. 1998. and L. Moskowitz.. 2000. Second Edition. Second Edition. 2003.. Pyke and R.W.F.Bibliography 55 Bibliography Askin. Montgomery. NJ. Scheduling. 1994.S. Inventory Management and Production Planning and Scheduling. McGraw Hill. Spearman. Prentice Hall. and Inventory Control. 1974. Elsayed. John Wiley & Sons. Silver. Decision Systems for Inventory Management and Production Planning. Foundations of Inventory Management. Zipkin. 1997.P. Third Edition. Wright. Prentice Hall. 1997. E. and T. Production: Planning. Production and Operations Analysis. L. Fourth Edition. and C. Operations Research Techniques for Management. Prentice Hall. P. W. Tersine. Prentice Hall.A. Irwin.A. NJ. R.A. Nahmis.. Englewood Cliffs. Krajewski. D. Prentice Hall.. Taylor III. E. Third Edition. Standridge. 1993. NJ. and E. and G. Fourth Edition.P. Operations Research in Production Planning. New York. New York.