fu.

€ q* *B* Es
=-=€.*8"*8,=

. i\TS 10.6 ORDER-qUANTITY,

-
REORDERPOINT MODEL
: INOMIC ORDER WITH PROBABILISTIC
{NTITY (EOQ MODEL DEMAND
-, How-Much-to-Order The How-N{uch-to-Order Decision
i:cision The When-to-Order Decision
- - \\ hen-to-Order Decision
: : -:itivity AnalYsis for the EOQ 10.7 PERIODIC REVIEW MODEL
-1. Jel
WITH PROBABILISTIC
DEMAND
: -:: Solution of the EOQ
',1,.del More ComPlex Periodic Review
':nary of the EOQ Model Models
-
--.sumPtions APPENDIX 10.1
DEVELOPMENT OFTHE
: ,r\OMIC PRODUCTION
OPTIMAL ORDERQUANTITY
. . SIZEMODEL
(o*) FORMULA FOR THE EOQ
- .. Cost Model
MODEL
. -,rmic Production Lot Size
APPENDIX 10.2
ENTORY MODEL WITH DEVELOPMENT OFTHE
;
-:.\NED SHORTAGES oPrlMAL LOT SIZE (0*)
, *-:1\TITY DISCOUNTS FOR FORMI.ILAFORTHE
EOQMODEL PRODUCTIONLOT Sr7F
: ' :LE-PERIOD MODEL
:\TOitY MODEL WITH
. - 3-\BILISTIC DEMAND
: : -in Marcus
- iu.ide Car Rental
:,
. -:: Chopter l0 lnveniory Models

Inventory refers to idle goods or materials held by an organization for use som---r .r

the future. Items carried in inventory include raw materials, purchased parts, comp_ ::
subassemblies, work-in-process, finished goods, and supplies. Two primary reaSrtri.: --l
nizations stock inventory are: (1) to take advantage of economies-of-scale that exi:: --.,,
the fixed cost of ordering items, and (2) to buffer against uncertainty in customer ,:.-* .,

or disruptions in supply. Even though inventory serves an important and esseni:. - r:.

the expense associated with financing and maintaining inventories is a substanria- - --

the cost of doing business. In large organizations, the cost associated with inven: : rrL

run into the millions of dollars.

In applications involving inventory, managers must answer two important que:

j. How much should be ordered when the inventory is replenished?

1. When should the inventory be reptenished?

Virtually every business uses some sort of inventory management model or Sl:-:- r

address the preceding questions. Hewlett-Packard works with its retailers ro::: -rl
termine the retailer's inventory replenishment strategies for the printers and t'-::' *iri
products. IBM developed inventory management policies for a range of n---. r

: .a itn entory pt'ocedure tronic parts that are used in IBM plants as well as sold to a number of outside - _ :
:: ,tribed for the drugstore ers. The Management Science in Action, Inventory Management at CVS Ccr, -_. ,

:':;;rsrn ls discussed in
describes an inventory system used to determine order quantities in the :--_ - nl
:::ail in Section 10.7.
industry.
The purpose of this chapter is to show how quantitative models c&r :: r

making the how-much-to-order and when-to-order inventory decisions. \i-e ,; .i

consider deterministic inventory models, for which we assume that the rate c: -: : *uL.
for the item is constant or nearly constant. Later we will consider probabilis:. - ' rl
tory models, for which the demand for the item fluctuates and can be descn:,: -
in probabi listic terms.

INVENTORY MANAGEMENT AT CVS CORPORATION*

CVS is one of 'the largest drugstore chains in the order quantity by counting the numL'e:
United States. The primary inventory management of the product on the shelf and subrr.;---- :r,,
area in the drugstore involves the numerous basic quantity from the replenishment level. .\ : -:- -r r
products that are carried in inventory on an everyday program determines the replenishmen: - ..*r"
basis. For these items, the most important issue is for each itemin each individual store. :r,:. rrr

the replenishment quentity or order size each time each store's movement rather than on th; - - r,:,r,,
an order is placed. In most drugstore chains, basic movement. To minimize stock-outs the ::- rr-: i

products are ordered under a periodic review inven- ment quantity is set equal to the store's '.::- , *r
tory system, with the r-eview period being one week. demand 01: movement for the product.
The weekly review system uses electronic
ordering equipment that scans an order label af-
fixed to the shelf directly below each item. Among *Based on information provided by B::
-. :
other information on the label is the item's replen- (The inventory system described uas .-- .,.
ishment level or order-to-quantity. The store em- implemented in the CVS stores formerlr r-- ,

ployee placing the order determines the weekly SupeRX.)

45q
I O. I Economic Order Quontity (EOA) Model

10.1 HCCIN$h{IC ORI}I{}I QUANTITY iFrSQi MOI}EI-

for an item
The economic order quantity (EOO model is applic:ablewhen the demand
shows a constant, or neafly constant, rate and when the entire
quantity ordered arrives in
that the same
inventory at one point in time. The constant demand rate assumption means
units every day,
number of units is taken from inventory each period of time such as 5
:: associated with 25 units every week, 100 units every four-week period, and so on'
::rg and maintaining To illustrate the EOQ model, let us consider the situation faced by the
R&B Beverage
," products' From a
Ir?" -.. is larger than
Company. R&B Beverage is a distributor of beer, wine,'and soft drink
:: ople think. Models retail stores with
main warehouse locatedln Columbus, Ohio, R&B supplies nearly 1000
H: ;he ones presented.
of the company's
:iapter can be used beverage products. The beer inventory, which constitutes about 407o
F :.,:ip cost-ffictive total inventory, averages approximately 50,000 cases. with an average cost per case of ap-
P : -. manaSement
ltt- proximately $S, RAg"rtimates the value of its beer inventory to be $400,000.
The warehouse manager decided to conduct a detailed study of the inventory costs
.:

F -r.

purpose of the study is

associated with Bub Beer, the number-one-selling R&B beer. The
to establish the how-much-to-order and the when-to-order decisions for
Bub Beer that will
I
manager
I
result in the lowest possible total cost. As the first step in the study, the warehouse
' :he most criticized obtained the following demand data for the past 10 weeks:
|
!r---ront of the EOQ
ln". :, rhe constont
fi:-: rate. Obviously,
Ir.:etwouldbe
lp.
- ,nriate
for items Week Demand (cases)
lfr ' .iel\'fluctuating and
demand rares. 1 2000
fc-,t 2025
' . .r. as rhis examPle 2
| 1950
h :he EOQ model -1

I:.,,idearealistic 4 2000
ln' : ttnation of the J ?l,0a
fu; order quantity 6 2050
| :,mand is relatively 1 2000
fr i ;nd occurs aI a 8 tg'l5
f' nrrtont rotr.
9 1900
I

10 2000
i

I
Total cases 20.000
I
Average cases per week 2000

Strictly speaking, these weekly demand figures do not show a constant demand rate'
However, given the relatively low variability exhibited by the weekly demand, inventory
planning with a constant demand rate of2000 cases per week appears acceptable. In practice,
you will find that the actual inventory situation seldom, if ever, satisfies the assumptions of
the model exactly. Thus, in any particular application, the manager must determine whether
the model assumptions a.re close enough to reality for the model to be useful. In this situation,
because demand varies from a low of 1900 cases to a high of 2100 cases, the assumption of
constant demand of 2000 cases per week appears to be a reasonable approximation.
The how-much-to-order decision involves selecting an order quantity that draws a com-
promise between (1) keeping small inventories and ordering frequently, and (2) keeping
iarge inventories and ordering infrequently. The first alternative can result in undesirably
high ordering costs, while the second alternative can result in undesirably high inventory
nolaing costs. To find an optimal compromise between these conflicting alternatives, let
iliryu
illilllll

4,-4* Chopter l0 lnventory Models

us consider a mathematical
moder that shows the
Aswithorherquantitative
*oJl:,:f:"^f:or,.' total cost as the sum
of the -: _ir:.rli

;,;;i:,::::;,y::!:#:;, ,-_. Io,oing "o.t, *. the costs associared wirh m,;hr";-;-^

.

,7,,,, po**ii,u inventory; tt

"r"
.ort, a"pend on the size ,r*.'11.T:ii11Tl*
or.. carrying a
€- :
';:';?i"';,:::,::,f
;,r i;,:Hff:j.1?T'Jf;tr,"i,u",tof"''ffiffi":;n3#ffifi*j;l1l:*,,
* u:"'its own money, it
theinvenroryhotiingcostwithnotb.i;g;;.;;;;:";il,,}T'?.'J1;111riencesanopportuni+C:.
cott with nor hein- ^Lr^ .-
the inventory horelinp
,"-"-^^L],-,-
"^perien""*
t;iiTi:r:;:;:";:,:Jil::3;IHj;tll;l*Xr.Tri"fl
reJbrs;;;;,;;:;;;,'' '
u'hich centage of the amount
frliF:ilffi 11",':ffi iU::.,:;.
costs.
costs. A number of or of capital urrurrr"
o*,.ffii1;i*"?""::llates
at an
orher i,"roirg its ""r, "-r."ri"iu, annua- ..
:..
r ""Jr, ffi::,:l:s-cost
::*ilxffi
hording
:T1l:,:;l;ffi;;:ffi
ro. tr,"'iie u.illr;";
t#"J1p"1ffi *,t,;:..;
,llil'iff
the value of its jnventory
The.cost "ott
of one case oi'uU g.".
,i,r'
rWilXiii' * 7Vo : 25Vo of rhe value or. T-.*_:- _ -.-l
h o I din g
ir-ig.
'tii t T#1iTxa";
l' r: :
I
::r :L",,';: ?i#
The next step in th
._e
l--^'rurve,Lury ror
:;"":,fr
rnvenrory analysis
;ffi
= $2.00.
is ro determir;;;r"d*;i.,*
" yearis 0.25($8) ":'
o

which is consiaerea _.
I
,,ji.h:.,,d ;;;;;::il'"-,Tjn[ HH:*:r,;,* ;;"^ "^.
;: o.er:--. : :
pavment'
t
tation,
tt
invoice
o.a..irg
verification."r.;.;;;;";:;'::'::'1c
receiving' and so on. F-o;
postage.
ntg ;LI";.t:"-;:
t.i.pl,. : -*:,rm

I
" "-
"-o,,-;nffi;1"''
tr':.:::#;#ft,.::"#ili:lffi,T#"'.:,i!",:?;1;ft
hour. the Iahar
u wage rate
fl,J;;"#n:_.:
and fringe benefit cosr fnr nr,"^r..
^^.,:^- :r^:n M#
H
cost, and demand inrormadon
-rrll! ?.,1#;:l ij*"'* or the EoQ model..After
are rhe rhree :- -
R&B probrem.
il1" atY:thow thev are used to a.u.lop*uo",Jooi.;'ri:;.= _-
we can
ov aennirg
o;r;.";;:""* ;;;;,"#, ,.
,t
"
uur.. o'r!?ii,]iffi ,1i#:lifi:L'|; ow-mu ch - to o,0". a".j
J. n .
-
orderi-ng costs.f;
tnventory r"r-ruu Beer
_,_ T:
size Q is received from
wilt have a *rr1':i 1Ovalue
of O unit. --
the c.nnr;o- D o n .,, -tt"' ',,,

received. Thus, assumin* another shipment oi : ,;

" "Jrr#;:Jj::".fte
rui:t;tit*Jiii-^.T"";Ttrl"ffi
period in qr"rtion.
rrrii
li;lll,rJlll:*Arii*l-
.'"'" (rrq(
invenr, -.
lIrE grapn Indlcates
an-average
i, o. tr,.,,jil,,?,, ;:ljffi,,H,*::ffff
;:::9,
oecllnes
au e becau s-e' ;;;
clines
r
;.. :.:rr

Figure l0.I
ro.l shows inven+^*, ,
,,r. ir*rao.;ffiH
strowsire at a constant rate
rare o\::
Lr\:- ---
'-- :; -| .

on' this o"n.* one order cvcre of 1::-:---

Most in,entorycost 9::r *,,'i'l#::::'l1[ |ffi,itlsinventory pattern i, ,t,nr. - -
lr tne uu.;;;."i;;'rvill-repeat' The complete ,**,
tnodets use an annuor
cost. dunng each cycle is
he of cycles is also yre."t "".."t. ir";il'
Thts. denund .rhorrrtr
YzQ, t',,.
p.(r,ressct! in
unit.t per year, The holding c<_lst can be
calculated
';:,:i;;:;:,:;u "sn'",1,, l:t:-if-' l",o,p .* o, murriprying th. usinc t inventory. rh,i
annuat rate. in invenrory ror rhe uu"iusT,1:.:':c"
,tut.a p.iiJo. ffi",ln"T::,[]!}?#,?: J:_ _ ,
beoneweek.onemonth.on.y.*.-o,;;;:;,,.".,';;;,;.il,::.
Hil:i:,", ,

tr* n"rn, onolysts Vpicoll refer io "fotor cost"

totor voriobre or totor moders fo
,ereu..ry n" t:: J;;" ;;;'#i:::::-:::l1Ti;i5:ii:;=-=-
" ,,u, c:: : I
order decision ore
considered rrxed or ill':
consiont ond ore not - -
included in ,r." r.a"rl
d*l
I O.l Economic Order Quontity (EOa| Model

$B$ilsffi*
"*S"! INVENTORY FOR BUB BEER

*o- Maximum lrwentorY

h
Average
t tlto Inventory

T
\\
Length of time required to
deplete an inventorY of P units

Tirne

FiSURH *S.? . INVENTORY PATTERN FoR THE EOQ INVENToRY MODEL

Inveotory is used at the constant

demand rate

rO
s
- *Average
t/2A Inventory

inventory models arc

industries and businesses is expressed as at annudl percentage, most
developed on at annual cost basis.
Let

/: annual holding cost rate

C : unit cost ofthe inventorY item
Ch : annual cost ofholding one unit in inventory

The annual cost of holding one unit in inventory is

{ t{i, } .}
Crr = /C
fa-

4S: Chopter 10 lnventory Models

Cy is the cost of holding

The general equation for the annual holding
one unit in inventory
for cost for the average inventory of yze ,
one year. Because smaller
as follows: --
ru

order quantities e will

result in lower inventory,
total annual holding cost
can be reduced by using ',,Annual o"
smaller order quantities.
holdiflS,cost = (,f,:;il? (^'*::L'
)
\ per uflrt /1
_ r ^,^
:;QCN
L'-

To complete the total cost model, we must

now include the annual ordering co::
goal is to express the annual ordering :r,r
cost in terms of the o.0", qruriity
tion is, How many orders will be placed during e. Thefirs: : _ .
the year? Let D denote the annual dc:-ru,
for the product. For R&B Beverage, D : 6;
r""tr;1ZOOO .ur", p", ,"ek) : l.* , i
cases per year. we know that by ordering Q units every time *"
C. fixed cost per
the place D / Q orders per year. If co is ttre
ord"., we wirl h..: l
order, is independent of coitit placing one order, the general equari: . - ,r.
the annual ordering cost is as follows:
the amount ordered. For a
given annual demand of
D units, the total annual
ordering cost can be
reduced by using larger
order quantities.

".fff;l"*: (;**)(:[:)
. (f)", :' j:

Thus, the total annual cost, denoted TC,

canbe expressed as follows:

Total Annual Annual

* ordenog
annual = holdiog
cost cost cost
TC=
{,air;i'[ci
Using the Bub Beerdata [Cr, : IC : (0.25)(gg) : $2, C" : $32, andD : 104.Cr-,
the total annual cost model is

,c:;e$2)+%*($:21 :,.ry
The development of the totar cost model goes
a rong way toward sorving the invenri,
problem' We now are able to express the
totlal annual u, a function ,1 y.,l* *r-ry,i*_;.
total cost "ort
be ordered' The development of a realistic
model is perhaps the most impou
part of the application of quantitative methods
to inventory ie"rrion making. Equar:.
(10'4) is the general total cost equation
for inventory situations for which the assumptic:,
of the economic order quantitymodel are valid.

ritilll
4e3
1O.l Economic Order Quoniity (EOa) Model

E Ee* E{*&}'-.&€ aleBt -s*- {} rder E'}**is&c}ee

cost for
Q that will minimize the total annuaT
The next step is to find the order quantit
y
: 8000' The total
BubBeer.Usingatrial-and-error^approach,wecancomputethetotalannualcostfor
a starting point, let usionsider Q
several possible order quantities. As
annual cost lbr Bub Beer is
3.328.000
rc:Q*- a
3.328.000
= 8000 * =g000
: $8416

A trial order quantity of 5000 gives

rC: 5ooo.'#P: $5666

Theresultsofseveralothertrialorderquantiti",u,",ho*ninTablel0.l.Itshowsthelow-
estcostsolutiontor:eauoutz000cases.Graphsoftheannualholdingandorderingcosts
10'3'
and total annual costs are shown in Figure
Theadvantageofthetrial-and-errorapproachisthatitisrathereasytodoandprovides
thetotalantrualcostforunu.u",orpossluteorderquantitydecisions.Inthiscase,the
disadvantage of
to be approximately 20d0 cases. The
l1:e EOQformula minimum cost order qo*iiry appears order quantity'
not provide the exact minimum cost
:trcrmines the oPtimal this approach, however, is tiiarit does
:,der quantiy b)' balancing RefertoFigurel0.3.Theminimumtotalcostorderquantityisdenotedbyanordersize
:;te annudl holding crtst and
ofQ*.Byusingdiff.erentiut"ut",to,,itcanbeshown(seeAppendix10.1)thatthevalueof
;lw unnuttl ordering cttst' cost is given by the formula
a{rnuminimizes the total annual

?DQ; { 1{i.5:i
Q*- Cn

order quantiry @OQ) formula'

This formula is referred to as the economic is
total annual cosi order quantity for Bub Beer
Using equarion f f O.ii it -inirno*
"
2(lo4go)2 :
.l
824 cases
Q*:

TOTAL COSTS FOR VARIOUS

E&*L€ 1*.t ANNUAL HOLDING, ORDERING, AND
ORDER QUANTITIES OF BUB BEER

A$gttr*E {lerst
Orderlng Total
{}t'iltr S*sst*3t-t- IIoIding
$ 666 $5666
5000 $s000
$ 832 $4832
4000 $4000
$1 109 $4109
3000 $3000
$1664 $3664
2000 $2000
$3328 $4328
1000 $1000

rliil
.t*4 Chopter 10 lnvenrory Models

FgSLJffig i*.;T ANNUAL HOLDING, ORDERING, AND TOTAL

COSTS FOR BUB BEE;

?
a
Annual Ilventory
L)
Holding Cost

Annual Ordering
Cost

I
I
I

3000 4
Order Quantity (ei

old"r_qrntitv
Problem 2 at the encl of the Lr^-1:: T for Bub Beer of 1824 in equatior (10.4) shows rhar
lnventory"tpolicy the mininr_-
chapter asks you to show has a total annual cost of $3649. Note
ances the holding and ordering costs.
that O* : I : __ .

that equal holding and Check for yourselfto see that these costs
ar. ._,
ordering costs is a pruperty
of the EOQ model. T'fu * E&'€Ee* ^ $et - *s-<*slE- E_.}*{: isi$sr
Now that we know how m,ch to order, we
Tlte reortLer point is want to address the question of w,hen t
t:xpressed in terms of To answer this question, we need to introduce
the concept of inventory posit:,-
inyentory positi on, the inventory position is def,ned as the amount ,

of inventory on t ura-flus the am _-

amount oJ inventory on inventory on order. The when-to-order
hand phts the Llmount on the
decision is
inventory position at which a new order
i, t".rn, oiu reorder p,,,:
"xpr.sred
.

order. Witlt short lead should be placed.

The manufacturer of Bub Beer guarantees
times, inventory pos ition a two-clay delivery
on any order pi.-.
is Ltsually the same as R&B Beverage. Hence. assuming R&B Beverage ,
operates 250 days per year. the _-
the inventory on hand. demand of 104'000 cases implies a da,y
demand of 104,000/2-50 = 416 case-s. T:
Hotvever, with long lead expect (2 days)(4r6 cases per day) : g32 _
cases of Bub to be sord during the
time s, inNentory position takes a new order to reach trre ian - r* _

nay be larger than warehouse. In inventory terminorogy,

delivery period is referre<i to as the read the r.,
intenton'on hand. time for a new order, and the g32_case :=-
anticipated during this period is referred
to as the lead-time demand. Thus, R&B ._
order a new shipment of Bub Beer from
the manufacturer when the invento^ ::
832 cases' For inventory systems using _

ihe constant demand ,ut. urru-ir#;" _-

2Actuolly'
Q* Irom equotion {10 5) is 'l 824 28, but becouse we connot
order froctionol coses of beer, o
is shown. This volue of
e* couse o few cents devioiion beiween the e- :
two costs. If q* is used ot its exoc. ,:
holding ond ordering costs ^ov
wili be exoctl;, the some.
d*5
I O.l Economic Order Quontity (EOAI Model

as the lead-time demand' For

these sYstems, the
lead time, the reorder point is the same
g-"*trf for the reorder point is as follows:
"-pt"ssion

irffi

where
r= reorder Point
d: dematdPer daY
m: leadtime for a new order in daYs

Thequestionofhowfrequentlytheorder.willbeplacedcannowbeanswered.Theperiod
p.Juioosly in equation (10'3)' we defined
as tire cycle ti-u.
r"r"Jio :
between orders is 1824 :
D l Q asthe number of otJ""
tf'u'*iffl" placed in a ye'ar' Thus' b l O: year'tO4'0001
If R&B places
will place ior Bub Beer each
57 is the number of ord; R&fi"rerage 4'39 working :
57 orders over 250 working davs'
it will orAe' appg*imatelv 1":y:::,1:'
for a cycle time3
days. Thus, the cycle ti*"-i* i':9 worti'g Aays' fne general expression
of 7 daYs is given bY

-ff ry-lf r--,r:I.:lii-il::I::--JiTll

250 750Q* {iil.l}
'- b/4"
r'.lr::::r.i':-i I -
D
::: ::'..'l r l

Nlodel
Sensitivity Analysis for th* E$Q
EventhoughSubstantialtimemayhavebeenSpentinarrivingatthecostperorder($32) good esti-
andthe holding r*'iiJ't. i"" ,t oorO ,"uiir"that these figures are at best
"rrr
mates.Thus,wemaywanttoconsidertrowmuctrttre.".o,n**d"dorderquantitywouldthe effects of
uno norJing costs. To determine
change with differen, several dif-
".ri*""0 "rdering order quantity
the recommerided
under
various cost scenarios, we can calculate
the total costbrder quantity for several
-ini*o*
ferent cost conditions. rru-r" io.z
shows
relativelv stable'
the tlble, the value of Q* appears
cost possibiliries. As d;;;;;fr";
evenwithsomevariationsinthecostestimates.Basedontheseresults,thebestorderquan.
properly, the total cost
170G-2000 cases. If operated
tiry for Bub Beer i, i, il" ,*ce, of per vear' we also note
irr";;;;;; should be close to $3400-$3800
for the Bub Beer quantity of 1824' For
implementing the calculated order
that little risk is associ""lr *lrr, (using cost estimates based
example, if R&B i*pr"*"nt,
o.d", quantity oi 1824 cases
-
on$32perorderaldZtvoannualholdingrate)buttheactualcostperorderturnsouttobe
$34andtheactualannualholdingrateturnsouttobe24vo,thenR&Bexperiencesonlya
annual cost'
6; i;;;;"t" ($3690-$3685) in the total
Fromtheprecedinganalysis,wewouldsay.,r,u..t,i'EoQmodelisinsensitivetosmall EoQ models
This insensitivity is a property of
variations or errors i, ;;;;;;rtimates. of ordering cost
have at least reasonable estimates
in general, wrrictr indicatl, itrri rr we true minimum cost
and holding cost, we of""'to obtain a good approximation of the

order quaniitY.
"*

iime would be given bv r = 30oQ* /

,**",o,expressionforcycletimeisbosedon250workingdoysperyeor.lfthefirmoperoled300workingdoys D'
**"1''"rti" *-kng'io;li[-tvtt"
per yeor ond wonted to "r
"';;;;;'''
+i;.5 Chopter 10 lnventory Models

TA*LS :*.I OPTIMAL ORDER QUANTITIES FOR SEVERAL COST POSSIBILITIES

F*sruihle {}ptimal Frcjectetl tetal

€g:v*:cti*r-.? E $ssEi!*e t3*"**r .Ammgral ilar-"{
Fl+i*Fi*:g {-"est p*r
{-l*si Qunrurity Using Ustng
i%i {}rder {#,*} Q* ltzt
Q =
u $30 1 803 $3461 $346:
24 34 1919 3685 369t-
26 30 1732 3603 360-
26 34 1844 3835 38i,f

Slxc*6 SqlE{eti$sx q*$- $Ea* {t{}* &.€+peieB

Inventory models such as the EOQ model are easily implemented with the
aid of .-,-,,,
sheets. The Excel EOQ worksheet for Bub Beer is shown in Figure 10.4.
The ,*o.,- ,,*
view of the formulas is on the left and the worksheet view of the l,alues is on the rigir:
on annual demand, ordering cost, annual inventory holding cost rate, cost per
- ,i
' unil. - ..;
i,g days per year, and lead time in days are input in cells S: to n8. The appropriai= : rrl
model formulas, which determine the optimal inventory policy, are placed in
B21. For example. cell BI3 computes the optimal economic order quantity
cells !
1g2-1.-: ;,L
:
M
$$SL:R* }*.4 WORKSHEET FOR THE BUB BEER EOQ INVENTORY MODEL

Vl'EBffi
EOQ
i

&*7 I
I O.l Economic Order Quontity (EOa) Model I

l
l

TASTH 10.3 THEEOQMODEL ASSUMPTIONS

at a coxstant rate'
1. Demand D is deterministic and occu$ increases bv Q units
2. The order quantity A; th; il;]or each tid;Jil;ilentory level
each tirne an order is received. ordered.
it ;o*n*'and dges not depend on the quantitry
3: The cost per ordo' C;, *i depend on the quantity ordered'
J*t'not
4. The purchase .or, p"i *it, C. is constan,
The total inventory
ne19o'.cr" is constant'
' The inventory h"ldi;;il
5.
illdft;;rii"p.,at"on
il;;ilft1e
uoitr cn and the size of the inventory'
6. Shortages such as *tf-""t orlackorders are not pormitted'
7. The leid time br an order is constailt
g. Theinvenrorypositio"nlJrerir"#d""o.i"*.:ly.Asaresult,anorderisplacedassoonas
point'
th; i;;;..y position reaches the reorder

is desired' one or
cost $3648.56' If sensitivity analysis
cell B16 computes the total annual or changes on the
more of the irprt autu ruio",
u" modified. The impact of any change
"*
policv will then appear in the worksheet'
;;;t;;ii.t
"nio.v
TheExcelworksheetinFigurel0.4isatemplatethatcanbeusedfortheEoQmodel.
presented in this
r^c; worksheets for ihe.other ir*tt"-11"-ys
This workshe"t ana smla.
chapter are available ir[ *tuor* iirt on the website that accompanies this text'
",

*fl th* H$Q &Eei{geE &ss*aratpti*ms

S*Eaaxaaar-Y

Tousetheoptimalorderquantityandreorderpointmodeldescribedinthissection,an
system op:lates. The EOQ model
analyst musr muf." ur.oirpiio* ui*t t o* theinventory
withitseconomicorderquantityformula.isbasedonsomespecificassumptionsaboutis provided in
of the assumptions foithis model
-i.iilly review the R&B invento$y;;. A ,umma.y to ensure
Table 10.3. Before ,rirg ir," EOQ
formula, carefully review these.assumptions
-, o.f the are not
If the assumptions
system being analyzed.
' :: before that they are applicable io the inventory
model'
. :-. tctual i"utonuUf", seek a different inventory
.-:, int'entory VarioustypesofinventorySystemsareusedinpractice,-andtheinventorymodelspre.
. ..: later in
sentedinthefollowingsectionsalteroneormoreorth"soQmodelassumptionsshown with different
- ',f ore or change, a different inventory model
, """tptions of in Table 10.3. when ihe assumptions
necessary'
optimal operating policies becomes

NOTES AND COMMENTS

the maxi-
lead+ime inventory drops to 2592 cases when
i.' With relatively long lead times' the .u* irr"ntoty level is 1824? The key is to re-
ana ttre resulting reorder point r de- in
a"tnunO
exceed g^' ;;; ftrat the reorder point is expressed
termined by equation (10'6), may which equals cases
lf this condition occurs' at least one order wlll
i*t of inventory i'on position
;or-fr*A' + cases the way'" Thus' to inter-
U" orttt*airg when a new order
is placed' For
r : 2592-realize that 2592total cases
:
has a lead time Dret the
;;;i;, ur.,irn" that Bub Beer
of d : 432
'*itt o..* when there we 1824 cases on the
way
;i;':6 days. with a daily demand
if.o* u p..rious order; and 768 on-hand' So' the
(10'6) shows th-at t:^ltoto"' place another order
point *ouUbe r = dm = 6 x 432
"ur"r, "qouiion : 2592 cases' -oJ.r tiu,.. that we should

Q* =.1824 *t.n ttt" on-hand inventory level is 768 cases'

fuoi" tirut this reorder point dxceeds
That is, because the lead time is so
long (6 days)'
to the maximum inven-
which also corresponds of units before the
At first glance' this *" fr"r"a place an order Q
tory-*- (see Figure 10'1)'
last order of O units has even arrived!
;;;; i*p"tsibli-how can we order when
__".-

4*E Chopter 'l

0 lnveniory Models

LA.2 B;{:{}H{}e,E-€{l E}Rt}SEl{l'f'A{}N E {}T SEF.*I &€{}E}gg,

The inventory model presented in this section is similar to the EOQ model ir t .r drrl
attempting to determine how much we should order and when the order shou-: :r -
The inventory model in this We again assume a constant demand rate. However, instead of assuming Li-:' t.-:r
section alters assumption 2 arrives in a shipment of size Qx, as in the EOQ model, we assume that units ?t; : ----1
of the EOQ model inventory at a constant rate over several days or several weeks. The constant srpryn
(see Table 10.3). The
assumption implies that the same.number of unrls is supplied to inventory ea;: :tri
assamption conceming
the arrival of Q units each time (e.g., 10 units every day or 50 units every week). This model is designec -'-r r".';Mltur
time an order is received tion situations for which, once an order is placed, production begins and a cons:=:r I
is changed to a constant of units is added to inventory eachday until the production run has been com;.:::
production supply rate.
If we have a production system that produces 50 units per day and we dec::. mu{lt
ule 10 days of production, we have a 50(10) : 500-unit production lot size. Iir= [lm
the number of units in an order. In general, if we let Q indicate the production . :i
approach to the inventory decisions is similar to the EOQ model; that is, we bui:: , I
and ordering cost model that expresses the total cost as a function of the produc-:,, hm
Then we attempt to find the production lot size that minimizes the total cost.
One other condition that should be mentioned at this time is that the modei : -:
to situations where the production rate is greater than the demand rate; the prr'-, .
tem must be able to satisfy demand. For instance, if the constant demand rate :. - r.rrrlllltrll

per day, the production rate must be at least 400 units per day to satisfy demar - ,ffiyit
During the production run, demand reduces the inventory while producl ,- l;:irTl tr,,if.ril

inventory. Because we assume that the production rate exceeds the demani -; *
day during a production run we produce more units than are demanded. Thus. ::,. n.q{ drffi
,,C{f
production causes a gradual inventory buildup during the production perio,J ;-r*y: ffid rlttil
production run is completed, the continuing demand causes the inventory to STa . *-j (*
This model differs from the cline until a new production run.is started. The inventory pattern for this sysre- ,tfl
EOQ morlel in that a setup in Figure 10.5. {il
cost replaces the ordering As in the EOQ model, we are now dealing with two costs, the holding c.-..r il ,] lnfil
cost, and the saw-tooth ,ifrttiil
ordering cost. Here the holding cost is identical to the definition in the EOQ :. ,:".
inyentory pattern shown in ilffi
Figure 10.5 dffirs.from the the interpretation of the ordering cost is slightly different. In fact, in a produ:-, r ,{tfl
inventory pattern shown in ation the ordering cost is more correctly refened to as the production setup c':r'x.
Figure 10.2. cost, which includes labor, material, and lost production costs incurred while :-,1 I

S$qJRS }*.$ INVENTORY PATTERN FOR THE PRODUCTION LOT SIZE

INVENTORY MODEL

*-Maximum
h

Time
44q
1O.2 Economic Produciion Lot Size Model
run
the production system for operation,
is a flxed cost that occurs for every production
regardless of the production lot size'

!r
iu;
Total il*s{ &$e*deE

u Letusbeginbuildingtheproduction-lotsizemodelbywritingtheholdingcostinterms for average

approach is to develop an expression
Ult, of the production lot size b.Again, the
h, inventoryandthenestablistrtt,"t'otai,gcostsassociatedwiththeaverageinventory.We
Sl' ,." u or"-y.ar time period and an annual cost.for the model'
u, IntheEoQmodeltheaverageinventoryisone-halfthemaximuminventory,or|zQ. rate
Figure 10.5 shows th"t f";;;;uction
lot size model, a constant inventory buildup
m
occursduringtheproductionrun,andaconstantinventorydepletionrateoccursduringthe
rhus, rhe average inventory will be
one-half the maximum inven-
*ii ;;;;;;,f;H;;, produciion lot size Q does not go into inventory
ffi, tory. However, in this;;;;;rtenithe
h at one point in time, and thus the
inventory never reaches a leve1 of Q units'
inventory' let
m To show how we can compute the maximum
m,
d: dailr demand rate
p: daily Production rate
l*
nit''
,: number of daYs for a Production run
u&
be larger than d,the daily inventory
buildup rate dur-
Because we are assuming thatp will p - d units in
. :.-int, the logic of I
ing the production ptrasJis p d' If *"
run-production for r days and place
t ' .tctton lot size will be (p - d)r' From
ilmr
end of the production run
t, " : easiertofollow inventory each day, tt it'i't"'V at the is the maxi-
rffi " u,,rre inventory at the end;f the production
,.",t run also
.. . : :ailY demand rate Figure 10.5 we
n' .-- : jaily- Production "un
ntr mum inventorY. Thus,
. ' -'!ovever, when the
frt .. -.-'.iul cost model is
mffir ' ;.:t develoPed, we Maximurr inventory : {P -'d)t f itl"8:
-.-ind that inPuts to
n- ,:rl be exPressed in
ffim
'the annual demand IfweknowweareproducingaproductionlotsizeofQunitsu::u,u,productionrate
Mr - , ;td the annual : pt, artdttre iength of the production run ' must be
l&r . , '.in rqte P- of p units, then Q
tu
ml ,''Q '''' i i {}.-i.ri
,,=,,l:.7;' day8

Thus,

Maxirnnm inventory + \P -'d)t = {B :

=(, t:u}p {ii,i.liii

the maximum inventory' is given by

The average inventory, which is one-half

* ii{.}.tli
Average inventorv
|(' - l\9
47fi
Chopter 10 lnventory Models

per-unirholding cost
Xf*,:Hl"al of c1,. the generar equadon
for annuar hording cost
:

If D is the annual deman!{o1,r,"

p.oauola{i *ffiil;;;ffi
cost, *rri"rr,*ri, ,i"i,r""
ffi.firr#ffilserup "r,h"
urr-;;o;;;;i*.or, in the E,_,-

M
+$;; i 1t:

,.,ixxo,il,'lfr 3$",*.,T:itr#y;i.#:",i.,#;;.;;;;;";;;;;_ 11

D
d=
250
Now let P denote the
annual r^- ^rthe
nrnrtrrnri^^ for
'"nual production
Then
day. product if the product
were produc;_

P :250p and ,: P
. 250
Thus,a

4:
p
DJ2so
D
p/2so - F
Therefore, we can write
the total annual cost
model as follows:
-t"-:i|", 1"1:-.a.--ir:-"::rr,
I ,."'"1.t,,,,-,,.-t,:li_,:..,.,,,_,.,,,.-, i,,I.i.i.t*::;"'**f,.1,,.1-::;,.---::;-:itrr-:*:i_:_rr-::**_,___

ii*io'r= D/phordsregordressof
thenurirberof doysof
operorion;250doysisusedheremerery..
=