European Journal of Operational Research xxx (2013) xxx–xxx

Contents lists available at SciVerse ScienceDirect

European Journal of Operational Research

j o u r n a l h o m e p a g e : w w w . e l s ev i e r . c o m / l o c a t e / e j o r

Production, Manufacturing and Logistics

An economic order quantity model with a known price increase and

partial backordering
a b,⇑
Ata Allah Taleizadeh , David W. Pentico
a Department of Industrial Engineering, College of Engineering, University of Tehran, P.O. Box 11155/4563, Tehran, Iran b
Palumbo-Donahue School of Business, Duquesne University Pittsburgh, PA 15282-0180, USA

article info abstract

Article history: A constant unit purchase cost is one of the main assumptions in the classic Economic Order Quantity model. In practice,
Received 8 June 2012 suppliers sometimes face a known price increase. In this paper, we develop EOQ mod-els with a known price increase and
Accepted 8 February 2013 partial backordering under two different assumptions about when the increase will occur. We prove the concavity of the extra
Available online xxxx
profit functions for both scenarios if a special order is placed just before the price increases. A solution method is proposed
and numerical examples are presented.
Keywords:
EOQ 2013 Elsevier B.V. All rights reserved.
Partial backordering
Known price increase

1. Introduction and literature review order which is placed before or at time the price increase is to be effective.

When a supplier announces either a temporary reduction or a permanent
increase in the unit purchasing cost of an item, the buyer can generally
decrease his total purchasing cost by placing a larger-than-normal special
order. We analyze the second case in this paper. In addition to the decision
about whether and by how much to increase the order quantity, a problem that
has been studied by a number of other researchers, we are including partial
backordering of demand during stockout periods. The decision problem, then,
is to determine, recognizing both the imminent price increase and partial
backordering, the optimal quantity to or-der before a price increase.
This topic without partial backordering is investigated in sev-eral texts
and articles on inventory control and management, including Naddor (1966),
Brown (1967), Tersine (1967), Brown (1982), Tersine and Grasso (1978),
Silver et al. (1988), Markowski
(1986), and Gupta and Goel (1989). The problem was first formu-lated by
Naddor (1966). Brown (1967) developed a model that ap-pears to be different
from Naddor’s (1966). However, Brown (1982) has shown that there is no
significant difference between the two proposed models. They both assumed
that the buyer has an oppor-tunity at the end of the current EOQ cycle to
make a purchase at the current price and there will be a price increase for
future or-ders. Otherwise, the usual assumptions of the basic economic order
quantity model are made. Goyal and Bhatt (1988) assumed that n purchase
orders of equal size are placed prior to an (n + 1)st special
⇑ Corresponding author. Tel.: +1 412 221 6608.
E-mail addresses: Ata.taleizadeh@gmail.com (A.A. Taleizadeh), pentico@duq.edu (D.W.
Pentico).
Taylor and Bradley (1985) relaxed Naddor and Brown’s timing
assumption and assumed the price increase does not coincide with the end of
a regular cycle. Lev and Soyster (1979) developed an EOQ model with a
price increase and a finite planning horizon. Lev et al. (1981) studied an EOQ
model in which one or more of the cost parameters or demand will change in
future. They as-sumed that any change in the costs is likely to affect the
demand rate. Kingsman and Boussofiane (1989) investigated an inventory
control system in which the times between price increases follow a
probability distribution function. Markowski (1990) investigated two different
scenarios. The first one is the Special Order Strategy (SOS) in which a special
order will be placed and, when the inven-tory again reaches zero, the buyer
returns to an EOQ policy for all following orders. In the second one, which is
the EOQ Strategy (EOQS), the buyer forgoes the special order and continues
using an EOQ after modification for the price increase. He provided
expressions for the actual total cost of both scenarios for any time of interest
and showed that the choice of time horizon affects the choice of optimal
strategy. Yanasse (1990) developed an EOQ mod-el with an anticipated price
increase in which, in order to deter-mine the optimal order quantity, he used
the criterion of minimizing the maximum error in terms of cumulative costs.
Lev and Weiss (1990) developed a model which divides the planning horizon
H into two parts including a closed interval [0, T] and a half open interval (T,
H]. For the first part the unit purchasing cost is C and for the second one it is
increased to C + C0. Moreover, they as-sumed that the fixed ordering cost and
unit holding cost are differ-ent in the two intervals. The aim of their research
was to determine the numbers and sizes of orders during both finite periods.
Erel

0377-2217/$ - see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.ejor.2013.02.014

Please cite this article in press as: Taleizadeh, A.A., Pentico, D.W. An economic order quantity model with a known price increase and partial backordering. European Journal of
Operational Research (2013), http://dx.doi.org/10.1016/j.ejor.2013.02.014
2 A.A. Taleizadeh, D.W. Pentico / European Journal of Operational Research xxx (2013) xxx–xxx
(1992) investigated the effects of continuous changes in the unit purchasing
cost and holding cost on the optimal order quantity and annual cost. Tersine
(1996) developed an EPQ model with an announced price increase in which
shortages are permitted and constant fraction of produced items will be
defective. Abad (2006) used an EOQ model in a supply chain model
including a producer, a vendor and an end user in which the producer consid-
ers both temporary reduction and increase in unit purchasing cost separately
and the buyer places a special order in both situations. After a specific time
the vendor will increase his unit selling price to the end user and the demand
rate will be influenced by this decision. In Abad’s model the special order
quantity and increased selling price of the vendor are both decision variables.
Huang et al. (2003) developed an EOQ model with an infinite planning
horizon and a single announced price increase, with an option of placing a
special order just before the price increases. They extended previ-ous work in
which it was assumed that the special order is an inte-gral multiple of the new
EOQ quantity. They assumed that there is a single instantaneous price
increase and did not restrict the spe-cial order to be an integer multiple of the
EOQ quantity. They used the Cesaro limit of a savings function to determine
the optimal spe-cial order quantity. Then Lim and Rodrigues (2005) pointed
out that their savings function was not Cesaro summable and their solution
method was not correct. Shah (1998) developed a dis-crete-time stochastic
inventory control model for perishable items when the vendor announces a
price increase of units at some fu-ture time. In this model a constant fraction
of on-hand inventory deteriorates and shortages are not permitted. Hsu and
Yu (2011) studied an EOQ model with imperfect quality items under an an-
nounced price increase. They assumed that the defectives are screened out by
a 100% inspection process and the defectives can be sold as a single batch at
the end of the inspection process. Ghosh (2003) developed an EOQ model
with full backordering in which an announced price increase is considered.
Depending on the length of the intervening period between the announcement
date and the effective date of price revision, two different models are
proposed.
Our second addition to the basic EOQ model is the inclusion of partial
backordering of demand during stockout periods. The first model for the basic
EOQ with partial backordering was by Mont-gomery et al. (1973). Additional
models that address the basic EOQ with partial backordering at a constant
rate, which is the assumption we will use here, include Rosenberg (1979),
Park (1982), Wee (1989), San José et al. (2005), Pentico and Drake (2009),
and Taleizadeh et al. (2012, 2013). While these models dif-fer in their choice
of decision variables and, to some extent, the cost structures used, the only
way in which their assumptions dif-fer from those of the basic EOQ model is
that they allow stockouts, with only a fraction of the demand during the
stockout period being backordered. Many other authors have extended these
mod-els to include such considerations as a backordering rate that in-creases
as the replenishment time gets closer, deteriorating inventory, and demand
that varies depending on the selling price, the inventory level, or the passage
of time. A comprehensive sur-vey of this research may be found in Pentico
and Drake (2011). In this paper we will use the decision variables and cost
structure in Pentico and Drake (2009).
The paper most similar to the model we develop here is by Sharma
(2009), which presents models for both a temporary price cut and a permanent
price increase. In essence, he adds an immi-nent permanent price increase to
the composite EPQ model in Sharma and Sadiwala (1997), which adds partial
backordering at a constant rate and less-than-perfect-yield to the EPQ model.
Shar-ma’s model differs from ours in a number of ways, some of which are
significant. When we discuss the assumptions of our model we will describe
how they differ from those in Sharma (2009).
Please cite this article in press as: Taleizadeh, A.A., Pentico, D.W. An economic order quantity model with a known price increase and partial backordering. European Journal of
Operational Research (2013), http://dx.doi.org/10.1016/j.ejor.2013.02.014
2. Problem definition and assumptions
Consider a situation in which a supplier announces that a price increase
for an item will take place at or before a buyer’s next scheduled ordering
time. A logical response is to order additional units (place a special order) to
take advantage of the lower (cur-rent) price prior to or at the regular
replenishment time. Thus the manager must decide whether to place a special
order and, if so, he must determine the size of the special order. If the price of
an item will increase by an amount C 0 at a specific time, then the unit cost
before that time will still be C and after that time will be C k = C + C0. When
the unit cost is C, the economic order quantity is:
¼ rffiffiffiffiffiffiffiffiffiffi ðÞ
Q 2AD 1
iC
After the increase, the EOQ will decrease to:
sffiffiffiffiffiffiffiffiffiffiffi
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffiffi
2AD C
Qk ¼ i C C0 ¼ Q C C0 ð2Þ
ð þ Þ þ
So the purchaser either orders a special quantity Q S to take advantage of
the lower price or ignores this opportunity and uses Q k for all future orders. If
a special order is to be placed, then the purchaser must determine the optimal
value of QS, the size of the next order, after which all future orders will be of
size Qk.
Our assumptions are basically the same as those used in Talei-zadeh et al.
(2012), modified to reflect that the buyer is interested in taking advantage of
the current lower price before it increases rather than buying at a sale price. In
the following we discuss, where relevant, how they differ from those in
Sharma (2009):
1. Shortages are allowed and a constant fraction b of the unsat-isfied
demand will be backordered.
2. There are different costs per unit for backorders, which is a cost per
unit per period, and lost sales, which is a cost per unit and includes the
lost profit on the lost sale. Sharma assumes that these two costs are the
same and that they are both a cost per unit per period, so in his model
there is a single cost for shortages based on the average shortage level.
3. All orders placed after time t1 will be at the new, higher cost per unit.
4. Unlike Sharma, we assume that yield is 100%, so we ignore the costs
of inspection. We also ignore the costs of in-coming transit, although
those costs and inspection costs can be included in the fixed and
variable costs of an order.
5. The order is paid for at the time of receipt, so we do not con-sider, as
Sharma does, the holding cost of in-transit inventory.
6. The fixed ordering and unit backordering costs for both the regular
and special orders are the same.
7. The holding cost per unit will increase after the price change to reflect
the higher unit cost.
8. We assume that the unit selling price will not change, so the cost of a
unit of lost sales will remain the same for the spe-cial order at the
current price and will decrease to reflect the higher unit purchase price
for future orders, which is not the case in Sharma’s model since he
does not recognize the lost sale cost as being different from the
backordering cost.
9. Related to #6 and #8 is another significant difference between our
model and Sharma’s: the treatment of the max-imum stockout and
backorder levels for normal and special orders. Since we recognize
that the cost of a lost sale is higher for a special order, due to the lower
unit cost, than it is for a post-change order, we allow the maximum
stock-
 A.A. Taleizadeh, D.W. Pentico / European Journal of Operational Research xxx (2013) xxx–xxx 3

out levels to differ for the two situations and to differ from what they Tk Length of inventory cycle at the new price
are for current orders at the current price. Since he does not recognize TS Length of a one-time special inventory cycle at the
the difference between the per-unit backordering and lost sale costs, current price
Sharma assumes that the maximum stockout level is the same for both (⁄) Indicates the optimal value
a regular order and a special order. Other variables
ATP The annual total profit without change in the unit
10. We assume, as in a basic EOQ model, that the entire order is delivered purchasing cost
all at the same time. Sharma assumes that there is a finite ATC The annual total cost without change in the unit
replenishment rate. purchasing cost
CTP The cyclic total profit without change in the unit
3. Model development and assumptions purchasing cost
CTC The cyclic total cost without change in the unit
We introduce the parameters and the variables of the model in Section 3.1. purchasing cost
We next review some of the previous work on the two basic components of CTPn The cyclic total profit when a normal order is placed at
our model: the known price increase problem without backordering and the the increased unit purchasing cost
EOQ with partial backordering in Sections 3.2 and 3.3 respectively. Our new CTPS The cyclic total profit when a special order is placed
model is developed in Section 3.4. In Section 4 we develop the equations for before increase in purchasing cost
determining whether to make the special purchase at the existing price and, if
so, what the order quantity should be. In Section 5 we show how to modify
the model developed in Sections 3.4 and 4 to allow for the delivery of the
item at a finite rate, the assumption that differ-entiates the EPQ model from In Section 3.2, we review the model to determine the order size for a
the EOQ model we are focusing on here. special order before an increase in price without backordering and in Section
3.3 the basic EOQ model with partial backordering is reviewed; these are the
two bases for our new model. Finally our development of a new EOQ model
with partial backordering when there is an increased purchasing cost at a
3.1. Notation
specific time is shown on Section 3.4.

The following notations are used to model the problem.

3.2. Tersine’s EOQ model with announced price increase – no shortage
Parameters:
A Fixed order cost To obtain the special order quantity when an increase in the pur-chasing
b Fraction of shortage that will be backordered price will occur at a known time at which the inventory level will be q,
CCurrent unit purchase cost of an item Tersine (1994) proposed the following total cost function – translated into our
0
C Scheduled increase in the purchase cost decision variables – if a special order is placed:
Ck Unit purchase cost after the increase in price
hDT2
D Demand quantity of product per period S

g Goodwill loss for a unit of lost sales: g = p

0
(P C) TCS ¼ A þ CðDTS qÞ þ 2 ð3Þ
h Current unit holding cost per unit per period, h = iC If no special order is placed, the total cost over an interval of length T S
h
k Unit holding cost per unit per period after the increase
that starts at the time the price increases, during which the first (at time t 2)
in price, hk = iCk
and all subsequent orders cost Ck = C + C0 per unit, is (see Fig. 1):
i Inventory carrying cost rate, percent per period p
Backorder cost per unit per period
p0 Lost sale cost per unit at the current price: q hq2
0
p = g + (P C) TCk ¼ CkðDTS qÞ þ hkDTk TS D þ 2D ð4Þ
pk0 Lost sale cost per unit after the increase in price. To determine the optimal size of a one-time special order, the difference
pk0 ¼ g þ ðP C kÞ in total costs, TCk TCS, which is the extra profit, must be maximized. Tersine
P Unit selling price determined the optimal size of the special order. Since we will be modeling
Decision variables:
and solving the problem using T S, the cycle length, rather than Q S, the order
bMaximum shortage level for a normal order at the current price
quantity, we translate Tersine’s order-quantity result into the length of a
bk Maximum shortage level for an order at the new price special order cy-cle if backordering is not allowed:
bS Maximum shortage level for a one-time special order at rffiffiffiffi
¼hþ C ¼h þ ðÞ
the current price CT
ffi

F Percentage of demand that will be filled from stock for a T C0 k k C0 T Ck 5

S
normal order at the current price C
Fk Percentage of demand that will be filled from stock for Using TCk TCS and replacing TS in Eq. (3) by Eq. (5), the obtained
an order at the new price extra profit is:
FS Percentage of demand that will be filled from stock for a !
one-time special order at the current price T 2 þ hq2
S

Q Order quantity for a normal order at the current price

Qk Order quantity for an order at the new price TCk TCS ¼ A T 1 2D hqTS: ð6Þ
QS Order quantity for a one-time special order at the
3.3. Pentico and Drake’s model for an EOQ model with partial back-
current price
T Length of a normal inventory cycle at the current price ordering

To obtain the economic order and shortage quantities when there are both
backordered and lost sales, when b 100

Please cite this article in press as: Taleizadeh, A.A., Pentico, D.W. An economic order quantity model with a known price increase and partial backordering. European Journal of
Operational Research (2013), http://dx.doi.org/10.1016/j.ejor.2013.02.014
4 A.A. Taleizadeh, D.W. Pentico / European Journal of Operational Research xxx (2013) xxx–xxx

(0 < b < 1) percent of customers whose demands are not satisfied immediately
T 2A h þ bp ½ð1 bÞp0 2 11
will wait to receive their backorders, Pentico and Drake (2009) proposed the ¼ hD b bh Þ
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
following cyclic total profit and cost functions (see Fig. 2): ffiffiffiffipp ð
F ð1 bÞp0 þ bpT 12
CTP ¼ ðP CÞDT ) ¼ ðh þ bpÞT Þ
2 þ bp ð1 2 Þ
0 ð1 bÞð1 FÞDT
(
Aþ þp p 2AhD ð13Þ
ð
hDT2F2
b1 ¼ 1 Dp0

ffiffiffiffiffiffiffiffiffiffiffi
D F 2T2 ffi

ð7Þ It should be noted that at unit price Ck we will have:

2 2 2 2
CTC A hDT F bpDð1 FÞ T 1 1 F DT 8
0
Tk 2A hk þ bp ð1 bÞpk0 2
14
hD h
¼ þ 2 þ 2 þp ð bÞð Þ ð Þ
¼
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
The average annual total profit and cost functions are:
ffiffiffikbpbkp ð Þ
ATP ¼ ðP( CÞD ) ð1 bÞpk0 þ bpTk
2 þ bpDð12 FÞ T þ p 0 ð1 bÞð1 FÞD ð9Þ F 15
Tþ

A hDTF 2
2
k ¼ ðhk þ bpÞTk ð Þ
bk ¼1 2AhkD
Dpk 0
ð16Þ
pffiffiffiffiffiffiffiffiffiffiffiffi
ffiffi

ATC A hDTF2 bpDð1 FÞ2T 1 1 FD 10

In the next section, the EOQ with partial backordering and a known price
0

¼T þ 2 þ 2 þpð bÞð Þ ð Þ increase is modeled.

The values of T and F that minimize ATC, only if b is at least as large as a
critical value b1 given by Eq. (13), are: 3.4. New model: EOQ with partial backordering and known price
increases

Suppose that the supplier of an item that is subject to partial backordering

of shortages announces a price increase that will oc-cur before its next regular
I ordering time. In this situation a logical response is to order additional units (a
special order) to take advantage of the current lower price. As pointed out by
DTS Silver et al. (1988), the structure of this decision problem (whether to place a
special order at a current lower price that is available for a limited time and, if
so how much to order) is very similar to the problem of whether to take
advantage of a temporary sale price, the problem analyzed in Taleizadeh et al.
(2012). Given the similarity of the solution models and solution process we
D will use here to those used in that paper, we will restrict the amount of detail
we will show in our analysis, referencing the relevant analyses in that paper.

DFT
D
DTk Two possible cases may occur: the price increase for the item occurs (1)
q when there is still inventory or (2) when there is a short-age. Thus we need to
have two scenarios relative to the inventory position:

t
t1 t2
T Scenario 1: The new, higher price starts during the positive inventory
T k
T
S
interval of the current cycle.
Scenario 2: The new, higher price starts during the stockout interval of the
current cycle.
For each scenario there are two possible responses by the buyer:
Fig. 1. EOQ model with increased price and no shortage.

Case 1: Place a one-time special order at the current price before the price
increases in order to gain the extra profit. Case 2: Do not place a special
order; continue using an EOQ pol-icy, switching from the current larger
EOQ based on the current lower price to a smaller EOQ based on the
increased price (see Eqs. (1) and (2)).

The profit functions will be modeled for both scenarios in com-bination

with both of the ordering possibilities cases. Then, for each scenario, the
difference between the two profit functions will be maximized to find the
optimum values for the special order quantity and shortage level.

3.4.1. Scenario 1: the price increases when there is no shortage

3.4.1.1. Case 1: a special order is placed. Assuming a replenishment lead time
Fig. 2. Simple EOQ model with partial backordering at a constant rate b. of zero, the price-change situation is depicted in Fig. 3.
Please cite this article in press as: Taleizadeh, A.A., Pentico, D.W. An economic order quantity model with a known price increase and partial backordering. European Journal of
Operational Research (2013), http://dx.doi.org/10.1016/j.ejor.2013.02.014
 A.A. Taleizadeh, D.W. Pentico / European Journal of Operational Research xxx (2013) xxx–xxx 5

Fig. 3. Inventory diagram when increased purchasing cost occurs before the stock position is exhausted.Inventory diagram when increased purchasing cost occurs before the stock position is
exhausted.

The price increase occurs before the stock position is exhausted, and there is hq2 D F 2T2
)
no opportunity for a regular replenishment. As shown in Fig. 3, a special CTPk ¼ Pq ( 2D þ þ gDð1
bp ð1 Þ
2 bÞð1 FÞT
order is placed at t 1 when the stock position is q (q P 0) units and the next
"
order will be placed at t3. If there is no special order, the next order will be þ ðP CkÞDbð1 FÞT þ PDTkðFk þ bð1 FkÞÞ
placed at t2. If the special order is placed, the profit function will be: 2 2
DF T bp D ð1 F T
2 2
( hk k k kÞ k

A þ CkDTkðFk þ bð1 FkÞÞ þ 2 þ 2 ð21Þ

CTPS ¼ PDTSðFS þ bð1 FSÞÞ þ gDð1 bÞð1 FkÞTk
)# TS
1 Tk FÞT D
q
hDT2F2 Þ
bp D ð1 FS 2T2
S S S

( )
A þ CðDTSFS qÞ þ 2 þ 2 FkÞTk ð17Þ Again using an approach similar to that used in Appendices A and B in
Taleizadeh et al. (2012), we can show that Eq. (21) can be simpli-fied as:
þgð1 bÞð1 FSÞDTS þ bCkD½ð1 FSÞTS 1
A h DF2T k
Using an approach similar to that used in Appendices A and B of Taleizadeh " 0
k k

et al. (2012), we can show that Eq. (17) can be simplified to:
CTPk ¼ ðP CkÞD ( Tk þ pk FkÞ þ 2

bpDð1 FkÞ2Tk
hDT2F2 T2 )# Dð1
" bÞð1
D F 2
þ 2 TS P CkÞD kÞ k)#
( S S bp ð1 SÞ S
b ð1
CTPS ¼ A Cq þ 2 þ 2 þ ð p0 P þ CÞDTS (
T k þ pk0Dð1 bÞð1 FkÞ þ
hk
2k k þ p 2
2
T
A DF2T D F
( bp ð1 2
) 1 FÞT þ D þ Pq 2D þ 2 Þ T
p0DFSTS p0 C0Þbð1 FSÞDTS CkbDð1 FkÞTk ð18Þ q hq2 D F 2
Note that, based on Fig. 3, the special order quantity and max-imum
backorder level for this case will be:

Q S ¼ DFSTS q ð19Þ )
þgDð1 bÞð1 FÞT þ ðP CkÞDbð1 FÞT ð22Þ
bS ¼ ð1 FSÞDTS ð20Þ
Using an approach similar to that used in Appendix C of Talei-zadeh et al.
(2012), the equation for the difference between the to-tal profit for the interval
3.4.1.2. Case 2: a special order is not placed. In this case the pur-chaser will
of length TS if a special order is placed (Case 1), given by Eq. (18), and the
not place a special order and will continue using a ‘‘standard’’ EOQ which,
total profit for an interval of the same length if a special order is not placed
instead of being based on the current unit cost C will be based on the new unit
(Case 2), given by Eq. (22), is:
cost Ck = C + C0. The total profit over the interval of length T S if a special
order is not placed is:
G1 ¼ w1TS þ w2FSTS w3FS2TS2 þ w4FSTS2 w5TS2 þ w6 1 ð23Þ
Please cite this article in press as: Taleizadeh, A.A., Pentico, D.W. An economic order quantity model with a known price increase and partial backordering. European Journal of
Operational Research (2013), http://dx.doi.org/10.1016/j.ejor.2013.02.014
6 A.A. Taleizadeh, D.W. Pentico / European Journal of Operational Research xxx (2013) xxx–xxx

where
A hkDFk2Tk bpDð1 FkÞ2Tk 0 D1 F ð24Þ
þ
w1 ¼ Tk 2 þ 2 pk ð bÞ k
ð25Þ
w2 ¼ p0 bpk0 D
ðh þ bpÞD ð26Þ
w3 ¼ 2
w4 ¼ bpD ð27Þ

bpD ð28Þ
w5 ¼ 2
2
A DF T D F 2 T q
w " ( 0 hk
6 1 ¼ ðP Ck ÞD Tk þ pk Dð1 bÞð1 FkÞ þ 2 k
k þ bp ð1 2 k
Þ
k )# ð1 FÞTþD
hq2 D
2
F T 2 A þ Cq
( bp )
Pq þ 2D þ ð12 Þ þ gDð1 bÞð1 FÞT þ CkbDð1 FkÞTk
ð29Þ
P CkÞDbð1 FÞT

In the next section we will find a unique solution that mini-mizes G 1 in CTPS ¼ PDTSðFS þ bð1 FSÞÞ þ Pbq
Eq. (23) for the decision about whether to place a spe-cial order and, if so, for hDT2F2 D F T
2 2
(
Þ
S S bp ð1 S

what quantity.
S

A þ CðDTSFS þ bqÞ þ 2 þ 2
)
3.4.2. Scenario 2: the price increases when there is a shortage þ gð1 bÞð1 FSÞDTS þ bCkD½ð1 FSÞTS 1 FkÞTk ð30Þ
3.4.2.1. Case 1: a special order is placed. Assuming a replenishment lead time
of zero, this price-change situation is depicted in Fig. 4. The price increase
occurs after the stock position is exhausted, and there is no opportunity for a Again using an approach similar to that in Appendices A and B in Taleizadeh
regular replenishment. As shown in Fig. 4, a special order is placed at t 1 when et al. (2012), we can show that Eq. (30) can be simplified to:
the stock position is q(q < 0) units, the backorder level is bq units, and the
next order will be placed at t3. If there is no special order, the next order will hDT2F2 D F 2T2

p
S S S

be placed at t2. If the special order is placed, the profit function will be: CTPS ¼ (A P CÞbq þ 2 þ b ð1 2 SÞ

þ ðp0 P þ CÞDTS p0DFSTS p0bDð1 FSÞTS

)
bCkDð1 FkÞTk þ C0Dð1 FSÞTS ð31Þ

Fig. 4. Inventory diagram when increased purchasing cost occurs after the stock position is exhausted.

Please cite this article in press as: Taleizadeh, A.A., Pentico, D.W. An economic order quantity model with a known price increase and partial backordering. European Journal of
Operational Research (2013), http://dx.doi.org/10.1016/j.ejor.2013.02.014
 A.A. Taleizadeh, D.W. Pentico / European Journal of Operational Research xxx (2013) xxx–xxx 7

Note that, based on Fig. 4, the special order quantity and max-imum In the next section we will find a unique solution that mini-mizes G2 in
backorder levels for this case will be: Eq. (35) for the decision about whether to place a spe-cial order and, if so, for
what quantity.
Q S ¼ DFSTS bq ð32Þ
bS ¼ ð1 FSÞDTS ð33Þ 4. Solution method
3.4.2.2. Case 2: a special order is not placed. In this case the purchaser will
not place a special order and will continue using a ‘‘standard’’ EOQ which, To justify ordering a special quantity before increasing purchas-ing price,
instead of being based on the current unit purchasing cost C, will be based on the profit from ordering the optimal special quantity must be greater than the
profit from using regular EOQ order quan-tities for the same amount of time
the new unit cost Ck = C + C0. The total profit
) TS (CTPS > CTPk). To maximize the extra profit of both scenarios, both Eqs.
over the interval of length TS if a special order is not placed is: bÞq
CTPk ¼ ( ð 2 Þ þ gð1 bÞð1 FÞDT 2D gð1 (23) and (36) should be maximized. However, note that finding the optimal
bpD 1 F 2T2 bpq2 value of G1 or G2, whichever is relevant for the particular scenario being
faced, is not sufficient; that value must be positive to justify plac-ing a special
" order at the sale price.
þ ðP CkÞDbð1 FÞT þ PDTkðFk þ bð1 FkÞÞ
We note that the coefficients of the variable terms – those that include F S
( DF 2T2 bp D F T2
2

A þ CkDTkðFk þ bð1 FkÞÞ þ

hk

2
k k

þ
ð1

2
kÞ k
or TS or both – in G1 and G2 are identical. The only coef-ficients that differ
are the constant terms: w6 1 in G1 and w6 2 in G2. As we will show in the
q
TS 1 FÞTþ D following, the G1 and G2 functions are concave and we can, therefore, find
)#
þ gDð1 bÞð1 FkÞTk Tk ð34Þ the optimal values for FS and TS by simultaneously solving the equations
Again using an approach similar to that in Appendices A and B in Talei-zadeh found by setting the two first partial derivatives equal to 0. Since the
et al. (2012), we can show that Eq. (27) can be simplified as: expressions for the partial derivatives of G1 and G2 depend solely on the
A variable terms, not the constants w6 1 and w6 2, the partial derivative
0
CTPk ¼ ðP h
CkÞD b ð1
Tk þ pk Dð1 Þ
bÞð1 FkÞ C0Dð1 bÞð1
)#
FkÞ equations for the two scenarios are identical and, therefore, the equations for
k 2k kþ p 2 k k TS
the optimal values of FS and TS for the two scenarios are the same. However,
þ

DF2T D F 2T
we note that, although the optimal values of F S and TS are the same for the
two scenarios, the optimal order quantity will not be the same since
A determining that must also rec-ognize the starting inventory position, and the
h P CkÞD T
k þ pk0Dð1 bÞð1 FkÞ C0Dð1 bÞð1 FkÞ optimal values of G1 and G2 based on those identical solutions will be
þ k 2k b ð1
kþ p 2 k
Þ )#
k ð1 FÞT D
different because of the different constant terms. We note that this is the same
DF2T D F 2T q result that Taleizadeh et al. (2012) found for the problem of determining the
optimal ordering strategy when faced with a temporary price reduction.
bpDð1 FÞ2T2 Because the optimal values of FS and TS are identical for the two scenarios,
we will only show the derivation for Scenario 1.
( 2 þ gð1 bÞð1 FÞDT
bpq2
2D gð1 bÞq þ ðP CkÞDbð1 FÞT ð35Þ
Again using an approach similar to that used in Appendix C in Taleizadeh 4.1. Concavity
et al. (2012), the equation for the difference between the total profit for the
interval of length TS if a special order is placed (Case 1), given by Eq. (31) In order to obtain a closed form solution, we first need to show that the
and the total profit for an interval of the same length if a special order is not objective function is concave.
placed, (Case 2), given by Eq. (35), is:
w w Theorem 1. The objective function G1 is concave.
G2 ¼ 1TS þ F TS
2 S w3FS2TS2 þ w4FSTS2 w5TS2 þ w6 2 ð36Þ
where
A hkDFk2Tk bpDð1 FkÞ2Tk Proof. To prove that G1 is concave, the Hessian matrix equation (H) can be
w1 ¼ T k þ 2 þ 2 pk0Dð1 bÞFk ð37Þ used. Following the same logic as in Appendix D of Talei-zadeh et al. (2012),
w 0
2 ¼ p bpk0 D T
[FS, TS] H [FS, TS] is non-positive, so the function is concave.
ð38Þ
ðh þ bpÞD ð39Þ
w3 ¼ 2
w4 ¼ bpD ð40Þ
bpD ð41Þ
w5 ¼ 2
 2 2
@ G @ G
FS @FS
2
@FS@TS FS
½FS; TS H TS ¼ ½FS; TS 2 @G 2
@ G
2
3 TS
4@TS@FS @TS
2
5 ð43Þ
w2
A w w
0
w6 2 ¼ ðP CkÞD Tk þ pk Dð1 bÞð1 FkÞ ¼ 2TS 2w2FS þ 2 F
3 S 4 w5 < 0
Thus the local optimum for G1 is a global optimum. Following the same
þ
h
k 2k b
kþ p
ð1
2 k
Þ 2 T )#
k ð1FÞT D
approach used in Appendix E of Taleizadeh et al. (2012), taking the partial
DF2T D F q derivatives of G1 with respect to TS and FS respectively and setting them
þ ( ð 2 Þ þ gð1 bÞð1 FÞDT 2D equal to zero, gives:
bpD 1 F 2T2 bpq2 F w 2 þ w 4T S 44
S ¼ 2w3TS ð Þ
wF
g ð1 bÞq A þ ðP CÞbq þ bCkDð1 FkÞTk w1 2 S

h
) TS ¼ 2w3FS2 2w4FS þ 2w5 i ð45Þ
P CkÞDbð1 FÞT ð42Þ
Please cite this article in press as: Taleizadeh, A.A., Pentico, D.W. An economic order quantity model with a known price increase and partial backordering. European Journal of
Operational Research (2013), http://dx.doi.org/10.1016/j.ejor.2013.02.014
8 A.A. Taleizadeh, D.W. Pentico / European Journal of Operational Research xxx (2013) xxx–xxx

After some substitutions and algebra, we have; ders being filled from the R D units not being used to fill the new orders. An
implicit assumption is that none of those existing back-orders will be
F w1w5 þ w2w5 46 canceled; this implies that the number of backorders during the item’s receipt
¼
S w1w3 þ w2w5 ð Þ phase will not increase, but will decrease at a rate of R D. If a FIFO approach
w2 is being followed, the existing backorders will be filled before any new orders
TS ¼ 2w3FS w4 ð47Þ are filled. Their assumption was that the same fraction b of these new orders
will not wait, so, while the existing backorders will be decreased at a rate of
4.2. Solution feasibility R, new backorders will accumulate at a rate of bD until all the backorders that
existed at the time receipt began are filled. This distinction between LIFO and
For a solution to be feasible, TS must be positive and FS must be between FIFO as the approach to filling back-orders is important since it has
zero and one. Using the same basic approach as in Appen-dix F of Taleizadeh implications for how the equations for T and F are modified for the
et al. (2012), we can show that b must satisfy the following condition in order continuous delivery scenario of the EPQ with partial backordering.
to have FS 6 1:
pffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
For the FIFO approach to filling backorders, which is the ap-proach
2AhkD followed by Pentico et al. (2009), adapting the model in Section 3.4 simply
bP1 pk0D ¼ bk ð48Þ requires replacing h, hk, and p, in the equations in Sections 3.4 and 4 by h =
Note that this is exactly the same lower limit for b as given in Eq. (16). We h(1 D/R), hk by hk(1 D/R) and p by p(1 bD/R). As shown in Appendix C of
Pentico et al. (2009), the conversions needed for using the finite receipt assumption
can also show that in order to ensure that FS is non-nega-tive C0, the price
and the LIFO approach to filling backorders are more complicated.
increase, must meet a lower-bound requirement:
0
C0 > hpk ð1 bÞ h þ bpÞðhkTkFkÞ C 49
6. A potential application scenario
bp ¼ 1 ð Þ
Using the same approach as in Appendix G of Taleizadeh et al. (2012), we
In this section we describe a possible application of our pro-posed model
can show that TS is positive only if C0 meets the low-er-bound requirement
that is based on a situation faced by the stores of a drugstore chain in Iran. A
given by Eq. (49).
store’s supplier is a pharmaceutical com-pany; its customers are patients who
4.3. The solution process use one of their special drugs, such as those for diabetes, goiter, and blood
pressure, which are sold only by the stores in this chain. Decisions about
In order to determine whether a special order should be placed and, if so, when and how much of each drug to order are made independently by the
its size, the following steps should be done to determine individual stores.
F S ; TS ; Q S ; b S .
p If a patient goes to one of these drugstores to fill his prescription for one
1. Calculate bk 1 2Ahk D . of these special drugs but cannot buy it because the store is currently out of
pD

¼ ffiffiffiffiffiffiffiffiffiffi k
2A C
0
C ,F stock, he may do one of two things based on how much of it he still has. If he
0

S 1, has enough of the drug to wait until the drugstore replenishes its stock, then
2. If b P bk, go to Step 3. If b < bk, T ;TS ¼ h þ T k

hD
C
his order will be backordered. But if patient does not have enough of the drug
QS DTS q, bS 0. Stop. ¼ qffiffiffiffi qffiffiffiffi
¼ ¼ 2
to wait, he will go to another drugstore in the chain and will buy it from that
3. Calculate Tk 2A h k þbp ½ð1 bÞpk
0
and Fk
0
ð1 bÞpk þbpTk . one. In this situation the sale at the original store will be lost. If we as-sume,
h D bp bh p ¼ h b pÞ Tk which is the case, that a fraction of the patients can wait and the remaining
þ
¼
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffiffiffiffikk ð k fraction will go elsewhere, the shortage at the original store will be partially
h i backlogged, which is one of the assumptions of our model.
0
4. Calculate C1 hpk ð1 bÞ hþbpÞð h k Tk Fk Þ .
¼ bp
0 0
5. If C >C , go to Step 6. If C 6 C , T C0 T Ck , F 1,
1 1 S
¼hþ qffiffiffiffi
C S
¼ A second characteristic of the store’s situation is that the cost to the store
Q S ¼ DTS , bS ¼ 0. Stop. of the drug can increase at irregular times (i.e., not on a fixed schedule). Now
2
h F T
k k k bpð1 Fk Þ
2
Tk p0 1 bF p0 bp0 pb assume that the store’s manager is informed that the unit purchasing price of
Aþ 2 þ 2 k Þkþ Þ
this drug will increase by a certain amount at a specific time in the near
DTk ð k
6. FS
ð

¼ A h F T
k k k bpð1 Fk Þ Tk
2
pk0 ð1 bÞFk ðhþbpÞþðp
0
bpk
0
Þpb future. The manager may want to replenish his stock of this special drug even
DTk þ 2 þ 2

p0 bpk 0
: Go to Step 7: though his inventory level has not yet reached reorder point, which may be
positive or even when the inventory level is negative (i.e., there are already
TS ¼ ðh þ bpÞFS bp patients waiting to receive the drug). Being rational, he realizes that he should
7. If Scenario 1, QS ¼ DFSTS q, else QS ¼ DFS TS bq and determine how much he should order in advance of the cost increase to
bS ¼ ð1 FS ÞDTS . Stop. increase his profit. The known price increase issue is the second assumption
of our paper.
5. Adapting the EOQ model to allow a finite delivery rate
Because we do not have access to the relevant information about demand,
Although the normal situation is that the order for a purchased item will costs, and the backordering rates that would be necessary to use this example
be delivered from the supplier in one batch, it is possible that the order will be to illustrate the solution procedure, we will do that with the examples in
received in a continuous stream at a finite rate R that is greater than the rate D Section 7.
at which they are being used to fill orders. As shown in Pentico et al. (2009),
the decision model for F and T for the EOQ with partial backordering at a 7. Numerical examples
constant rate b, which is the basis for our model in Section 3.4, can be easily
adapted to allow for a finite receipt rate. To illustrate the application of the solution procedure given above, we will
use three numerical examples. The parameter val-ues common to all the
In making this adaptation, it is important to recognize the dif-ference examples are D = 500 units/year, A = $50/or-
explained in Pentico et al. (2009) between what they call the LIFO and FIFO
approaches to filling backorders. If a LIFO ap-proach is being followed, the
new customer orders, which arrive at a rate of D, are filled as they come in,
with the existing backor-
Please cite this article in press as: Taleizadeh, A.A., Pentico, D.W. An economic order quantity model with a known price increase and partial backordering. European Journal of
Operational Research (2013), http://dx.doi.org/10.1016/j.ejor.2013.02.014
 A.A. Taleizadeh, D.W. Pentico / European Journal of Operational Research xxx (2013) xxx–xxx 9

der, h = $3/unit/year, p = $1/unit/year, p0 = $5/unit lost, C = $10/ unit,C0 = Step 2. Since b = 0.9 is greater than bk = 0.7172 go to Step 3.
$2/unit, Ck = $12/unit, p0k ¼ $3/unit lost, hk = $3.6/unit/ year, P = $14/unit, Step 3. From Step 3 of Example 2, Tk ¼ 0:5 and Fk ¼ 0:3333.
and i = 0.3. Any additional parameters needed will be defined in the example. Step 4. From Step 4 of Example 1, C1 = 1.6.
Step 5. Since C0 = 2 is greater than C1 = 1.6, go to Step 6.
Step 6. From Step 6 of Example 1, Fs ¼ 0:7222 and TS ¼ 1:2.
Example 1 (Scenario 1). Let b = 0.5 and q = 50 units. Step 7.
pffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
QS ¼ DFSTS bq ¼ ð500Þð0:7222Þð1:2Þ 0:9Þð 50Þ ¼ 478:3333
2Ahk D ¼ 1 2ð50Þð3:6Þð500Þ ¼ 0:7172.
qffiffiffi bS ¼ ð1 FSÞDTS ¼ ð1 0:7222Þð500Þð1:2Þ ¼ 166:6667
Step 1. bk ¼ 1 pk0 D ð3Þð500Þ ¼ ffi ¼
Step 2. Since b = 0.5 is less than 2A
2 50 ¼

bk 0:7172; T hD

qffiffiffiffiffiffiffiffiffiffið
Þ ¼ The optimal extra profit using Eqs. (36)–(42) is 2209.3628, which is
ð Þ 0:2582,
3 500 positive, so the special order should be placed.
qffiffiffi ¼ Þqffiffiffi
¼ þ ffi 3 þð ffi ¼ ¼
Q
0
C T Ck 2 0:2582 12 0:9495. Then S
T
S h C 10

DTS q ¼ 500ð0:9495Þ 50 ¼ 424:7547, bs ¼ 0 and the extra 8. Conclusion

profit will be equal to 491.25.
We developed EOQ models for two scenarios with an an-nounced price
increase and partial backordering. We consider that the purchaser either
orders a special quantity QS to take advantage of the current, lower price or
Example 2 (Scenario 1). Let b = 0.9 and q = 50 units.
uses the new regular EOQ model’s or-der quantity Q k. Since ordering a
special quantity before the price increases requires that the profit from the
Step 1. From Step 1 of Example 1, bk = 0.7172. optimal special quantity must be greater than the profit from following the
Step 2. Since b = 0.9 is greater than bk = 0.7172, go to Step 3. Step 3 normal EOQ pol-icy, we maximized the difference between the profit
functions for

¼ sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

ffiffiffikbbk sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
¼ ¼
T 2A hk þ bp ð1 bÞpk 0
2 2ð50Þ ð3:6Þ þ ð0:9Þ1 ½ð1 0:9Þ3 2
0:5
k
hD h 3:6 500 0:9 1 0:9 3:6 1

Þ
p p ðð Þ ð Þ ð Þð Þð Þ
0
F ð1 bÞp þ bpTk ¼ ð1 0:9Þð3Þ þ ð0:9Þð1Þð0:5Þ ¼ 0:3333
k

k ¼ ð hk þ bpÞTk ð3:6 þ ð0:9Þð1ÞÞð0:5Þ

Step 4 the special and regular orders. The concavity of the maximizing
function is proved. Combining the ideas in Pentico and Drake’s
C hpk0ð1 bÞ h þ bpÞðhkTkFkÞ solution algorithm for the basic EOQ model with partial back
1 ¼ bp ordering and in Tersine’s approach to solving the price-increase
¼ ð3Þð3Þð1 0:9Þ 3 þ ð0:9Þð1ÞÞð3:6Þð0:5Þð0:3333Þ problem without backordering, a heuristic algorithm is proposed
0:9 1 to solve the model and numerical examples are presented to dem-
Þ
ð

ðÞ onstrate the performance of the algorithm. An advantage of using

¼ 1:6 our modeling approach, in which the decision variables are T, the
cycle length, and F, the fill rate, is that the optimal solution variable
Since C0 = 2 is greater than C1 = 1.6, go to Step 6. values are the same for both scenarios. Since the optimal order
Step 6 quantities and stockout levels can be easily determined from the

A hk Fk Tk
2
bpð1 Fk Þ Tk
2
0 1 F 0 0
DTk þ 2 þ 2 pk ð bÞ k þ p bpk pb
FS ¼ A
2
hk Fk Tk bp 1 Fk Tk
2
pk0ð1 bÞFk ðh þ bpÞ þ p0 bpk0 pb
DTk þ 2 þ ð 2 Þ

50 ð3:6Þð0:3333Þ ð0:5Þ
2
þ ð0:9Þð1Þð1 0:3333Þ ð0:5Þ
2
3Þð1 0:9Þð0:3333Þ þ ð5 0:9Þð3ÞÞ ð1Þð0:9Þ
ð500Þð0:5Þ þ 2 2 ¼ 0:7222
¼ 50 3:6 0:3333 2 0:5 0:9 1 1 0:3333 2 0:5

ð500Þð0:5Þ þ ð Þð 2 Þ ð Þ
þ ð Þð Þð 2 Þ ð Þ 3Þð1 0:9Þð0:3333Þ ð3 þ ð0:9Þð1ÞÞ þ ð5 0:9Þð3ÞÞð1Þð0:9Þ

T
p0 bpk
0
5 0:9ð3Þ 1:2
Example 3 (Scenario 2). Let b = 0.9 and q = 50 units.
S¼ ¼ ¼
ðh þ bpÞFS bp ð3 þ 0:9ð1ÞÞð0:3333Þ 0:9ð1Þ
Step 1. From Step 1 of Example 1, bk = 0.7172.
Step 7. Q S ¼ DFS TS q ¼ ð500Þð0:7222Þð1:2Þ 50 ¼ 383:3333
bS ¼ ð1 FSÞDTS ¼ ð1 0:7222Þð500Þð1:2Þ ¼ 166:6667

The optimal extra profit using Eqs. (23)–(29) is 2194.1128, which is

positive, so the special order should be placed.

optimal values of T and F, this makes the problem easier to solve than using
the more common approach of Tersine (1996) and Shar-ma (2009), in which
the decision variables are the order quantity and stockout level, which are not
the same for the two scenarios. For future work, adding a deterioration rate
can be suitable.
Please cite this article in press as: Taleizadeh, A.A., Pentico, D.W. An economic order quantity model with a known price increase and partial backordering. European Journal of
Operational Research (2013), http://dx.doi.org/10.1016/j.ejor.2013.02.014
References
Abad, P.L., 2006. Quantity restrictions and the reseller’s response to a temporary price reduction
or an announced price increase. Asia–Pacific Journal of Operational Research 23, 1–23.
Brown, R.G., 1967. Decision Rules for Inventory Management. Holt, Rinehart & Winston, New
York.
Brown, R.G., 1982. Advanced Service Parts Inventory Control, second ed., Materials
Management Systems, Norwich, Vt.
10 A.A. Taleizadeh, D.W. Pentico / European Journal of Operational Research xxx (2013) xxx–xxx
Erel, E., 1992. The effects of continuous price changes in the EOQ. Omega 20, 523– 527.
Ghosh, A.K., 2003. On some inventory models involving shortages under an announced price
increase. International Journal of Systems Science 34, 129– 137.
Goyal, S.K., Bhatt, S.K., 1988. A generalized lot size ordering policy for price increases.
Opsearch 25, 272–278.
Gupta, Y.P., Goel, S., 1989. Optimal ordering quantity for announced price increases.
International Journal of Systems Science 20, 2347–2350.
Hsu, W.K., Yu, H.F., 2011. An EOQ model with imperfective quality items under an
announced price increase. Journal of the Chinese Institute of Industrial Engineers 28, 34–
44.
Huang, W., Kulkarni, V.G., Swaminathan, J.M., 2003. Optimal EOQ for announced price
increases in infinite horizon. Operations Research 51, 336–339.
Kingsman, B.G., Boussofiane, A., 1989. Ordering and stockholding under price inflation when
prices increase in successive discrete jumps. Engineering Cost and Production Economics
17, 395–407.
Lev, B., Soyster, A.L., 1979. An inventory model with finite horizon and price changes. Journal
of the Operational Research Society 30, 43–53.
Lev, B., Weiss, H.J., 1990. Inventory models with cost changes. Operations Research 35, 53–
63.
Lev, B., Weiss, H.J., Soyster, L., 1981. Optimal ordering policies when anticipating parameter
changes in EOQ systems. Naval Research Logistics 28, 267–279.
Lim, A., Rodrigues, B., 2005. A note on the optimal EOQ for announced price increases in the
infinite horizon. Operations Research 53, 731–732.
Markowski, E., 1986. EOQ modification for future price increases. Journal of Purchasing and
Materials Management 22, 28–32.
Markowski, E.P., 1990. Criteria for evaluating purchase quantity decisions in response to future
price increases. European Journal of Operational Research 47, 364–370.
Montgomery, D.C., Bazara, M.S., Keswani, A.K., 1973. Inventory models with a mixture of
backorders and lost sales. Naval Research Logistics Quarterly 20, 255–263.
Naddor, E., 1966. Inventory Systems. John Wiley & Sons, New York.
Park, K.S., 1982. Inventory model with partial backorders. International Journal of Systems
Science 13, 1313–1317.
Pentico, D.W., Drake, M.J., 2009. The deterministic EOQ with partial backordering: a new
approach. European Journal of Operational Research 194, 102–113.
Pentico, D.W., Drake, M.J., 2011. A survey of deterministic models for the EOQ and EPQ with
partial backordering. European Journal of Operational Research 214, 1179–1198.
Please cite this article in press as: Taleizadeh, A.A., Pentico, D.W. An economic order quantity model with a known price increase and partial backordering. European Journal of
Operational Research (2013), http://dx.doi.org/10.1016/j.ejor.2013.02.014
Pentico, D.W., Drake, M.J., Toews, C., 2009. The deterministic EPQ with partial backordering:
a new approach. Omega 37, 624–636.
Rosenberg, D., 1979. A new analysis of a lot-size model with partial backlogging.
Naval Research Logistics Quarterly 26, 349–353.
San José, L.A., Sicilia, J., Garcia-Laguna, J., 2005. The lot size-reorder level inventory system
with customers impatience functions. Computers & Industrial Engineering 49, 349–362.
Shah, N.H., 1998. A discrete-time probabilistic inventory model for deteriorating items under a
known price increase. International Journal of Systems Science 29, 823–827.
Sharma, S., 2009. On price increases and temporary price reductions with partial backordering.
European Journal of Industrial Engineering 3, 70–89.
Sharma, S., Sadiwala, C.M., 1997. Effects of lost sales on composite lot sizing.
Computers & Industrial Engineering 32, 671–677.
Silver, E.A., Pyke, D.F., Peterson, R., 1988. Inventory Management and Production Planning
and Scheduling, third ed. John Wiley & Sons, New York.
Taleizadeh, A.A., Pentico, D.W., Jabalameli, M.S., Aryanezhad, M.B., 2013. An EOQ model
with partial delayed payment and partial backordering. Omega 41, 354-368.
Taleizadeh, A.A., Pentico, D.W., Aryanezhad, M.B., Ghoreyshi, S.M., 2012. An economic
order quantity model with partial backordering and a special sale price. European Journal
of Operational Research 221, 571–583.
Taylor, S.G., Bradley, C.E., 1985. Optimal ordering strategies for announced price increases.
Operations Research 33, 312–325.
Tersine, R.J., 1967. Materials Management and Inventory Systems. North Holland, New York.
Tersine, R.J., 1994. Principles of Inventory and Materials Management, fourth ed.
Prentice-Hall, New Jersey, pp. 113–115.
Tersine, R.J., 1996. Economic replenishment strategies for announced price increases. European
Journal of Operational Research 92, 266–280.
Tersine, R.J., Grasso, ET., 1978. Forward buying in response to announced price increases.
Journal of Purchasing & Materials Management 14, 20–22.
Wee, H.M., 1989. Optimal inventory policy with partial backordering. Optimal Control
Applications and Methods 10, 181–187.
Yanasse, H.H., 1990. EOQ systems: the case of an increase in purchase cost. Journal of the
Operational Research Society 41, 633–637.