Define useful capacity measures for a brewery

Capacity

Management

LEARNING OBJECTIVES After studying this chapter, you should be able to:

f0l rxntain the concept of capacity.

@ oescriUe how to compute and use capacity measures.

@ Descriue long-term capacity expansion strategies.

@ Desuue short-term capacity adjustrnent strategies.

1,O

@ rxplain the principles and logic of the Theory of constraints.

In April of 201.6 Tesla unveiled its low-cost Model 3 electric car. Over 325,000 potential buyers put up a deposit in the first week. While Tesla added new production capacity to increase production of its higher cost Model S by some :S,OO0 vehicles to about 100,000 in 2015, the Model 3 will bring some significant

capacity challenges. Jay Baron, CEO of the Center for Automotive Research in Ann Arbor observed: "They have to speed up production, and you're talking four or five times faster than they produce now. On top of all that you need more people. You need more stations on the assembly line."1

WHAT DO YOU THINK?

In a general sense,capacitg is alneasure ofthe capabil-

ity of a manufacturirrg or sewice systetn to perfbrm its

intended function. In practice, it is measured by the arrount of output tl-rat can be produced in a particular

time period-for exarnple, the number of harnburg- ers rnade during a weekday lunch hour or the nunrber

of patients that can be liandled during an emergency room shift. Having sufficient capacity to rneet cus- tomer dernand and provide l-righ levels of custotner

seru:ice is vital to a successful business and long-term

economic sustainability.

Capacity decisions cannot be taken llghtly and can

have profound impacts on business perfbrmance, as the

introduction suggests. For example' while Teslas factory

was optimized {br its rnodel S luxury velricle, the com-

pany will now require new assembly lines and scale up its

production capacity. This also means that they will need

to hire additional workers, engineers, and staffto supPort

1 98 PARTTHREE: Managing operations and Supply Chains

not only its value-cretrtion processes, but also suppor: processes such as scheduling, quality, training, and sup plier development. In addition, it will need to expand itr; network of retail and service centers :rcross tlle country

UNDERSTANDING CAPACITY

Capacity isthe copobility of o monufocturing or service resource such os o facility, process, workstation, or piece of equipment tc occomplish its purpose over a specified time period. Capercity can

be viewed in one o{'two ways:

1. as the maximum rzrte of output per unit of time; or 2. as units of resource availability.

For exarnple, tl're capacity of an automobile plant rnight be measured as the rrumber of automobiles capable of being

r-- I Capacity decisionsI I are often influenced bva'l

economies and diseconomies Iof sca1e. I

Capacity decisions can have a profound impact on

business oerformance.

produced per week and the capacity of a paper mill as the rrumber of tons it can produce per yeerr. As a resource availability rneaslrre, the capacity of a hospital would be rneasured by the nurnber of beds available, and the ca- pacity of "cloud" storage would be measured in gigabytes.

Operations managers rnust decide on the appro- priate levels of capacity to rreet current (sh<-rrt-term) and future (long-term) demand. Exhibit l0.l pr:ovides exarnples o{' such capacity decisions. Short-term ca- pacity decisions usually involve adjusting schedules or staf{ing levels. Longer-term decisions typically irrvolve major capital investments. To satisfy custorners in tlre long run, capacity must be at least as large as the av- erage demar.rd. However, dernand for many goods and services typically varies over time. A process may not be capable of rneeting peak demand at all times, resulting in either lost sales or customers who must wait until the good or service becomes available. At other peri- ods of time, capzrcity rnay exceed demar.rd, result- ing in idle processes or facilitjes, or buildups in pliysical inventories.

CHAPTER 10: Capacity Managemenr 199

Amount of overtime scheduled for the next weeK Number of emergencY room nurses on call during a

downtown festival weekend Number of call center workers to staff during the holiday season

Construction of a new manufactunng plant

Expanding the size and number of beds in a hospital Number of branch banks to establish in a new market terfltory

ro-ra Economies and Diseconomies

of Scale

Capacity decisions are often influenced by economit's

".,J dir""onomies of scale. Economies of Scale o'e

achieved when the averoge unit cost of a good or service decreos'ts

as the copocity and/or volume of throughput increases' For exan r-

ple, the design and construction cost per room of buikt-

ing a hotel decreases as the facility gets larger because

tlie {ixed cost is allocated over more rooms, resulting

in a lower unit room cost. This lends support to buikl-

ing larger {acilities with more capacity' Diseconc'- mies of scale occur when the overoge unit cost of the gocd

or service begins to increase as the copacity and/or volume rtf

throughput increases. In the hotel exarnple, as tlre nun'-

ber o{ roorns itr a hotel continues to increase, the ave'

age cost per unit begins to increase because of ltlrgt'r

airour-rts of o.rerllead and operating expenses required

by higher levels ol'such amenities as restaurants' parl -

ing, ,rnd recreational facilities' This suggests that some

optiroul amount of capacity exists-where costs are at a

,r]ini,l'tuto. The pressure on global automakers {br inr-

proving ""ono*i",

of scale is evident by Daimler and

Nirruriurrnourrcing plans to share the costs of develol'-

ing small-car technology, including engines' They will

sJap 3. I percent stakes in their companies with plans I o

generate $5.3 billion in savings over 5 years'2

As a single facility adds more and rnore goods ancVt'r

services to its portfolio, the facility can become tt o

" large and "unfbcused "

Clogged CoLlrts Court systems in most major cities are strained past

their capacity with case backlogs and long delays'

often resulting in freeing criminals before they go to

trial or incarcerating defendants who are subsequently

acquitted for long periods of time. Civil cases can be

stuck in court systems for years. Some court systems

are turning to technology and OM principles to

address these capacity issues. For example, better

computers and software for judges allow them to

scan conviction and sentencing documents right from

their courtrooms rather than recess to their chambers'

Police departments are also scanning reports, allowing

orosecutors faster access to information' lmproved

scheduling, such as blocking time for certain types of

cases that are backlogged and hiring part-time retired

judges, can help also. Can you think of other OM

practices that might help?3

SOLVED PROBLEM 10.1

An automobile transmission-assembly factory normally

operates two shifts per day, 5 days per week' During

each shift, 400 transmissions can be completed' What is

the capacitY of this factorY?

Capacity = (2 shifts/day)(5 days/week) x (400 transmissions/shift) x (4 weeks/month)

: 1 6,000 transmissions/month

At some point, disecono- mies of scale arise, ard unit costs increase b,:- cause dissimilar produ';t lines, processes, peoP e skills, and technology e <-

ist in the sarne facilitY. ln

tr)4ng to rnanage a lrrrl:e facility with too mar,Y

200 PARTTHREE: Managing operations and Supply Chains

A I

I

I

Units of demand

or capacrry

r'/'ca?acilY

_,-Average- demand Actual

*/ demand Minimum clemand

meet demand and satisfy customers. This is evident in Exliibit 10.2. Therefore, some amount of safety capacity (often colled the capacity cushion), defined os on omount of copocity reserved for unanticipoted events such os

demand surges, moterials shortages, and equipment breakdowns,

is normally planned into a process or {acility. In general, average safety capacity is de{ined by Equation 10.1.

Average safe$ capacity (Vo) :1007o - Average resource utilization (7o) [I0.1]

Note that Equation l0.I is based on average resource utilizations over some time period. For a factory, average safety capacity might be computed over ayear,whereas for an individual workstation, it might be updated monthly.

1O-2b capacity Measurement

/ work order rs o specifrcotion of work to be performed for o customer or o client. It generally includes the quantity to be produced, the process- ing requirements, and resources needed. W

shop) or services (e.9., a patient at a dentist's office

or room maintenance at a hotel). For any produc- tion situation, setup time can be a substantial part of total system capacity and therefore must be included in evaluating ca- pacity. Equation 10.2 pro- vides a general expression

-Time+

objectives and missions, key competitive priorities such as delivery quality, customization, and cost per{brmanc e can begin to deteriorate. This leads to the concept of- a focused {'actory.

A focused factory is a woy to ochieve economies of scale, without extensive investments in facilities and capocity, Ly

focusing on o norrow ronge of goods or services, target markct

segments, and/or dedicated processes to moximize efficiency onC

effectiveness. The {bcused factory argues to "divide an,l conquer" by adopting smaller, more focused facilities dedicated to (1) a few key products, (2) a specific tech- nology, (3) a certain pr:ocess design and capability, (4) r specific cornpetitive priority objective such as next-da r delivery and (5) particular market segrnents or custom- ers and associated volumes.

@ i^f lfftti,M

EAS u B EM ENT

Capacrty measures are used in many ways in long-terr planning and short-term management activities. For ex- ample, managers need to plan capacity contingencies fbr unanticipated demand and plan routine equipment and labor requirements. In this section we present several examples of how capacity measurements are used in OM.

ro-za Safety Capacity

The actr-ral utilization rates at most facilities are not planned to be 100 percent of efl'ective capacity. Unantic- ipated events such as equipment breakdowns, employee absences, or sudden short-term surges in demand will reduce the capability of planned capacity levels to

CHAPTER 10: Capacity Management 201

for evaluating the capacity required to meet a given pro-

duction volume for one work order, i'

Capacity required (C,) : Setup time (S') * [Processing time (P,) X Order size (Q,)1 [10.2]

where

C, : capacity requirements in units of tirne (e'g', minutes, hours, d.tys) for work order i

S. : setup or changeover tirne {br work order i as :r fixed atnount that does not vary with volume

p. : processing time for each unit of work order i (e.g', hours/part, minutes/transaction, etc' )

Q,: size of order i in nurnbers of units

If we sum the capacity requilelnents over all work orders, we can compute tlre total capacity required, using

Equation 10.3.

lC,: l[S, + (P x Q,)] lr0.3l

SOLVED PROBLEM !0.2

A typical dentist's office has a

complicated mix of dental procedures

and significant setup times; thus, it

is similar to a job shoP. SuPPose a

dentist works a 9-hour daY with one

hour for lunch and breaks' During

the first six months the practice is

open, he does all the work, including

cleaning and setting up for the next

dental procedure' Setup and processing times for three

procedures are shown in Exhibit 10'3' Also shown are the

number of appointments and demand for each type'

On a particular day, there are two scheduled

first appointments for single tooth crowns (see last

Manufacturing wclrk orders normally assrlme that

one setup is necessary for each work order, zrnd therefbre

the setup time is spread over the single work order quan-

tity. That is, setup time is independent of order size' Sorne services, sucl'r as hospital surgeries, may require a

new setup for ezrcl'r unit (i.e., the order size is one)' For

example, setup tirne for a surgery might include steril-

izing equipment, cleaning and disinfecting the surgical

suite, and preparing equipment {br the next procedure'

It is impoftant to understand these differences in goods-

producing and service-providing applications so work-

load is correctly cornPuted.

The concepts introduced in Equations 7.1 and 7'2 in

Chtrpter 7 can be used to calculate capacity:

Resources Used Utilization (U) :

Resources Available

Demand Rate

[7 1]

ln 01LI.AIUtilization (U) : Service Rate X Nurnber of Servers

column of Exhibit 10.3), one second

appointment for a single tooth crown,

four tooth-whitening appointments,

three first appointments for a partial

denture, and two third appointments

for a partial denture. ls there sufficient

capacity to Perform all the work?

We maY use Equation 10.3 to

compute the capacity requirements'

Exhibit 10.4 shows the results using the Excel Capacity

template. We see that a total of 610 minutes of work are

scheduled during a 480-minute workday' Therefore' there

is a capacity shortage of 130 minutes'The dentist will either

have to work two hours longer or reschedule some patients'

E

,i o

Sinole tooth crown ' .:t. ,,,

I

'1,. -; ' Tooth whitentng

Partial donture

4^4 I SL

2nd

1st

1st

2nd

3rcl

90

30

JU

30

20

JU

15

10

5

20

10

F

2

1

4

0

z

202 PARTTHREE: Managing Operations and Supply Chains

tooth crown lst

jcapafity Measuremcnt

JEnter data only In ycllow cells

OE

The templatc ls deslgnt d for up lo l0 work ordere.

1

J

1

8

9

10

dewntur€ lst

'-44i1}1.,ilT4:i.1."1:.:!!\::::4:.:.:\!:...4.i'j..i:]it:..'c;-i'::

From Equation 7.1, Resources Used : U X lte, sources Available. Wl-ren U < l, then Resources Used < Resources Available, so the actual capacity is dirninisJrei by the lack of full utilization of the resources. Thus, we can express capacity as

Resources Used (Capacity) : Utilization X [Service Rate X Number of Servers] [10.4]

This formula is used in Solved Problem 10.3.

11 iP.rtirl denture 3rd

*Assumes one setup time per dental procedure to se patient arrives.

From this analysis, we see that 21.3 percent of his total capacity is used to set up and change over from one dental procedure to the next. lf a dental assistant or technician is hired to do this work (assuming that this can be done off-line while the dentist continues

to work on other patients), revenue would increase by about 20 percent. lf setup times could be reduced by 50 percent, the total setup time would be 65 minutes instead of 130 minutes, and the capacity shortage would only be 65 minutes, requiring only one hour of overtime.

100 customers. This peak bour of demand happens two hours before every home football game. The average custorner purchase is

I burger (4-ounce hamburger or cheeseburger) 1 bag offrench fries (4 ounces)

1 soft drink (12 ounces)

Consequently, management would like to determine how rnany grills, deep fryers, and soft drink spouts are needed.

A 36 X 36-incli grill cooks 48 ounces of burg- ers every 10 minutes, and a single-basket deep fryer

cooks 2 pounds of french fries in 6 minutes, or 20 pounds per hour. Fi- nally, one soft drink spout dispenses 20 ounces of soft drink per minute, or I,200 ounces per hour. These effective capacity

E estirnates are based on 9,! the equipment manufac- ! turer's studies of actual E ,rr" under normal operat- 6 ing conditions.

ro-zc Using Capacity Measures lor 0perations Planning

Capacity needs rnust be trans-

lated into specific require- rnents for equipment and Iabor. To illustrate this. we present a simple ex- arnple. Fast Burger Inc. is building a new restau- rant near a college football

stadium. The restaurant will be open 16 hours per day, 360 days per year. Managers have concluded that the restaurant should the capacity to handle a peak hourly

have

demand of

up and clean the equipment and room before the next

CHAPTTR 1O: Capacity Management 203

SOLVED PROBLEM 10.3

The process for renewing a driver's license at the Archer County Courthouse is as follows. First, the clerk fills out the application; then the clerk takes the driver's picture; and finally a typist enters the info into the computer system and processes the new license. lt takes an average of five minutes to fillout an application, one minute to take a picture, and seven minutes to process the new license. lf there are two clerks and three typists, where will the bottleneck in the system be? How many drivers can be processed in

one hour if both the clerks and typists work at an 80 percent utilization rate?

The first two stages share the resources of the clerks, we assume the two clerks are cross-trained and pool

their efforts.The service rate for clerks : (60 min/ hour)/(5 * 1 min/driver) = 10 drivers/hour. Using Equations 7.2 or 10.4, we can find the capacity of the

{};!wfir3J:tiy': r f q9ij!{t?9i*, tlvl t f}x",rysqltrnlw.tr': ?F:*{.Y'?lsEa "wq -1 F".*

To determine the equipment needed to meet peal. l-rourly demand, Fast Burger must translate expecteC demand in terms of customers per hour into needs for grills, deep fiyers, and soft drink spouts. First note that

the peak hourly demand for burgers, french fries, and

soft drinks are as follow:

clerks in terms of the number of drivers per hour they can process:

Capacity Utilization (U) :

Service Rate x Number of Servers Caoacitv

0.8 : -;*orCapacity : l6drivers/hour10x2 For typists, the service rate is : 60/7 : 8.57 drivers/hour.

Again, using Equation 10.4, we have

Caoacitv 0.8 ::#orCapacity : (0.8)(2s.7)

8.57 X 3 : 20.6 drivers/hour

Because these processes occur in series, we see that

the capacity of the system is limited by the clerks (the

bottleneck), so the total process output is 16 drivers/hour.

{ir.r t1-rry:-rygflrs++dftt-f .414.:424 al#.el43 tri3':1{ql;fil4f?::gfe'5:st'ilr!:"'

Peak Hourly Product Demand (ounces) Rr rrnorq

French fries

Soft drinks

Finally, the nurnber of soft drink spouts needed to satisfy peak demand of 1,200 ounces is

Number of soft drink spouts needed : t.200/r.200 : 1.0

After revierving this analysis, the managers decided

to purchase two 36 X 36-inch grills. Grill safety capac- ity is 2.0 - 1.39 : 0.61 grills or 175J ozlhour [(0.61) x (48 oz/I0 min) x (60 min/hour), or about 44 harrburgels per hour. Management decided this excess safety capac-

ity was justilied to handle demand surges and grill break-

downs. With two grills the managers reduced their risk

of being unable to fill customer demand. If they installed two french fiyer rnachines, they would have 0.75 excess

machines, and that was thought to be wasteful' Horv- ever, they realized that if the one french fryer machine broke down they would not be able to cook enough fries,

so they decided to purchase two deep fiyers.

Management decided to go with a two-spout solt drink system. Although their analysis showed a need {br

only one soft drink spout, the managers wanted to provide

some safety capaciry primarily because they felt the peak

hourly demand for solt drinks might have been underes-

timated and customers tend to refill their drinks in this

self-service situation.

The average expected equipment utilizations fbr the two grills, two fryers, and two soft drink spouts are

as lbllows (refer to Equation 7.1):

400

400

1,200

Because the capacity of a grill is (48 o/10 minutes) (60 minutes/hour) : 288 ounceshour, the number of grills needed to sxisfy a peak hourly demand of 400 ounces of

burgers is

Nurnber of grills : 400/288 1.39 grills

To determine tlte number of single-basket deep fryers needed to r"neet a peak hourly demand of 400 ounces ol

french fries, we must first cornpute the hourly capacity

of the deep {ryer.

Capacity of deep fryer : (20 lb/hourX 16 oz-/Ib) : 320 ozlhour

Hence, the number of single-basket deep fiyers needed

is 400/320 : I.25. 204 PARTTHREE: Managing operations and Supply Chains

Grill utilization (U) : Resources used./Resources available

: r.39/2.0 : 69.5vo Fryer utilization (U) : Resources used/Resources

available : r.25/2.0: 62.5Vo

Soft drink spout utilization (U) : Resources used/ Resources availabl:

: 1.0/2.0:50.0Vo The managers of Fast Burger Inc. rnust also stalf the

new restaurant for peak dernand of 100 customers/ro,rr. Assume {iont-counter service personnel can take and irs- serrble orders at the service rate of 15 customers per hour and the talget labor utilization rate for this job is 85 percent. The nunrber offront-service counter people that should be assigned to this peak demand period crtn be found using Equation 10.4:

Utilization (U) : Capacity

Serwice rate X Number: of servers

Using the data for Fast Burger, we have

0.85 : (100 customer/hour: (15 customers/hour) X (Number of server i)

Solving tlis equatior-r for the number of'servers, we obtai r

(f 2.75)(Number of servers) : 100 Number of servers : 7.8

Given these capacity computations, Fast Burger mar - agement decides to assign eight people to the fronr- sewice counter during this peak demand period. Safety capacity is included in this decision in two ways. Firs , the target utilization labor rate is 85 percent, so there is a 15 percent sa{'ety capacity according to Equation 10.J. Second, eight people are on duty when 7.8 are needec, so there is a safety capacity of 0.2 people, The rnanage- ment at Fast Burger now has an equipment and labor capacity plan for this peak-demand period. Notice that equipment capacity, which is dif{icult to increase in thrr short-term, is liigh, whereas ltrbor js more easily changed. This equiprnent and labor capacity stratety must also bc coupled wiih good forecasting of demand-the subjec: of the next chapter.

LONG-TERM CAPACITY STRATEGIES

In developing a long-range capacity plan, a firm musl make a basic economic trade-off between the cost ol capacity and the opportunity cost of not having ade- quate capacity. Capacity costs include both the initial

investment in facilities and equipment and the an- nual cost of operating and maintaining them, rnuch of which are fixed costs. The cost of not having suf- ficient capacity is the opportunity loss incurred from lost sales and reduced market share. Having too rnuch capacity, particularly if demand falJs, can be devastat- ing. For example, International Paper recently closed its Franklin paper mill in Virginia. The mill had a ca- pacity of 600,000 tons per year, but demand had fallen 30 percent and was forecasted to continue falling, much of it a result of new technologies such as e-mail, Kindles and iPads, and electronic transactions.a Too little capacity, on the other hand, can squeeze profit rnargins or leave a firm vulnerable to competitors if it cannot satisfy customer orders.

Long-term capacity planning must be closely tied to the strategic direction of the organization-what prod- ucts and services it o{fers. For example, many goods and services are seasonal, resulting in unused capacity dur- ing the off-season. Many finns offer complementary goods and servi(el, which are goods ond services thot con be produced or delivered using the some resources ovailable to the

firm, but whose seosonal demand patterns ore out of phose with

eoch other. Complenrentary goods or services balance seasonal demand cycles and therefore use the excess capacity available, as illustrated in Exhibit 10.5. For in- stance, demand for lawnmowers peaks in the spring and summer; to balance manufacturing capacity, the producer might also produce leaf blowers and vacuums for the au- tumn season and snowblowers for the winter season.

10-3a Capacity Expansion

Capacity requirements are rarely static; changes in mar- kets and product lines and competition will eventually require a firm to either plan to increase or reduce long- tenn capacity. Such strategies require deterrnining the amnunt, timing, and fonn of capacity changes. To illus- trate capacity expansion decisions, let us make two as- surnptions: (1) capacity is added in "chunks," or discrete increments; and (2) demand is steadily increasing.

Four basic strategies {br expanding capacity over sorne {ixed time horizon are shown in Exhibit 10.6 (tliese

concepts can also be applied to capacity reduction):

1. One large capacity increase (Exhibit 10.6a)

2. Small capacity increases that match demand (Exhibit r0.6b)

CHAPTER 1O: Capacity Management 205

-rt---taf\.ii I I

A I

I

I

Units of oemano

or capacrty I I I I I I t t t

4.

-

Time

Srnall capacity increases that lead demand (Exhibit 10.6c)

Small capacity increases that lag demand (Exhibit I0.6d)

The strategy in Exhibit 10.6(a) involves one lar11e increase in capacity over a specified period. The adval- tage o{'one large capacity increase is that the fixed costs

ofconstruction and operating system setup needs to lre

Time

(a) One large capacitg increase

Time

(c) Small capacitu increases that lead demand

incurred only once, and thus the firm can allocate tbese costs over one large project. However, if aggregate de- mand exhibits steady grou,"th, the facility will be under- utilized. The alternative is to view capacity expansion incrementally as in Exhibit 10.6(b), (c), and (d).

Exhibit 10.6b illustrates the strategy of match- ing capacity additions witli demand as closely as pos- sible. This is o{ten called a capacity straddle strategy. When capacity is above the demand curve, the firm has

tme

(b) Small capacitg increases that metch domand

Time

Id) Smalt capacitg increase6 thet lag demand

206 PARTTHREE: Managing Operations and Supply Chains

Eriggs & Stratton is the world's largest producer of air-c rcled gasoline engines for outdoor power equipment. The company designs, manufactures, markets, and services these products for original equipment manufa,--turers worldwide. These engines are primarily aluminum alloy gasoline engines .anging from 3 through 25 horsepower. Briggs & Stratton is a leading designer, -nanufacturer, and marketer of portable generators, lawnmowers, snow throwers, pressure washers, and related accessories. lt also provides engines for marrufacturers of other small engine-driven equipment such as snowmobiles, go-kar:s, and jet skis.

The complementary and diverse original equipment markets for Briggs & Strafton engines allow factory managers to plan equipment and labor capacities and schedules in a much more stable operating environment. This helps minimize manufacluring costs, stabilize workforce levels, and even out volumes so that assembly lines can be used in a more efficient fasf,ion.s

E

9

?

excess capacity; when it is below, there is a shortage of capacity to meet dernand. In this situation, there will be short periods of over- and underutilization of resourct,s. Exhlbit 10.6(c) shows a capacity-expansion strategy with the goal of maintaining sufficient capacity to minimi re the chances of not meeting demand. Here, capacity ex- pansion leads or is ahead of'demand, and hence is calk,cl a capacity lead strategy. Because there is always exce;s capacity, safety captrcity to meet unexpected dernar d {iom large orders or new custorners is provided.

Finally, Exliibit 10.6(d) illustrates apolicy of acapa,-.- ity lag strategy that r:esults in constant capacity short- ages. Such a strategy waits until demand has increased to a point lvhere additional capaciq/ is necessary. It rc- quires less investment and provides for high capacity utilization and thus a higher rate of return on inves - ment. However, it mzry also reduce long-term pro{itabl - ity through overtirne, subcontracting, and productivily Iosses that occur as the firrn scrarnbles to satisfy demant[. In the long run, such a policy can lead to a permaner t loss of market position.

@ ilif^ll,ttxrvr 'APA'ITY

If short-terrn dernand is stable and sufficient capacib/ is available, then managing operations to ensure thai demand is satis{ied is generally easy. However, wher r demand fluctuates above and below average capacit.' levels as was illustrated in Exhibit 10.2. firms have twr'

basic choices. Fir:st, they can adjusi capacity to match the changes in demand by changing internal resources ancl capabilities. The second approach is to manage capacity by shifting and stimulating demand.

10-4a Managing Gapacity hy Adiusting Short-Term Capacig Levels

When short-ternr demand exceeds capacity, a firrn rnust temporarily increase its capacity, or it will be unable to meet all of the denrand. Similarly, if demand falls well below capacity, then idle resources reduce profits. Shor:t- term adjustments to capacity can be done in a variety of ways and are sumrnarized as follows:

limited by machine and equipment availability are rnore difficult to change in the short run because of high capital expense. However, leasing equipment as needed can accomplish this in a cost-effective manner. Another way is through irrnovative

l- I firms can adjust capacityI I to match changes inI

demand by changing internal resources and capabilities, or manage

r capacity by shifting and

I st imulat ing demand .

., I

CHAPTER 10: Capacity Management 207

partnership arrangements and capacity sharin g.

For example, a consoftium of several hospitals might be set up in which each hospital focuses ol a particular specialty and shares services.

SelI unused capacitq: Some firms might sell idle

capacity, such as computer storage space and computing capacif', to outside buyers and even t,l competitors. For example, hotels often develop partnership arrangements to accommodate their competitors' guests when they are overbooked.

Change labor capacity and schedules: Labor capa( ib/

can usually be manirged easily througli short- term changes in worklbrce levels and schedules.

Overtirne, extra shi{ts, temporary employees, and

outsourcing are common ways of increasing capacily.

Adjusting worHbrce schedules to better coincide

with demand patterns is another. Many quick-seruce restaurants employ large numbers of' part-time ernployees with varying work schedules.

Change labor sktll mix: H:iringthe right people who can learn quickly and adjust to changing job

requirernents and cross-training them to perform different tasks provides the flexibility to meet

fluctuating demand. In supermarkets, for examplt,, it is common for employees to work as cashiers during busy periods and to assist with stocking shelves during slow periods.

Shift uork to slack periods: Another strategr is to shift work to slack periods. For example, hote'

clerks prepare bills and perform other paperwor <

at night, when check-in and checkout activity is

light. Thls allows more time during the daytirne

hours to service customers, Manufacturers often

build up inventory during slack periods and hold

the goods for peak demand Periods'

r0-4b Managing Gapacity by Shifting

and Stimulating Demand

Some general approaches to influence custorners to sh ft derrrand frorn periods without adequate capacity !o periods with excess capacity, or to {ill times with excess capacity, include the following:

> Vary the price of goods or sensices: Price is the most powerful way to influence demand. For example,

hotels might offer cheaper rates on tle weekend; airlines might offer better prices for midweek

fliglits; a restaurant might cut its meal prices in hal{

after 9:00 p.m. In a similar {'ashion, manufacturers typically offer sales and rebates of overstocks to

208 PARTTHREE: Managing Operations and Supply Chains

stimulate demand, smooth proclucuon scnedules ano

staffing requirements, and reduce inventories.

Prorside custornerc with informafion: Many call centers, for exanrple, send notes to customers on their bills or provide an automated voice message

recornrnending the best times to call. Amusement

parks sucb as Disney World use signs and literature

informing customers when certain rides are extrernely busy.

Ado ertisin g and, p ro moti on: Afte r-holiday sales

are heavily advertised in an attempt to draw customers to periods of traditionally low dernand. Manufacturer or service coupons are strategically distributed to increase dernand during periods of low sales or excess capacity.

Add peripheral goods and/or sen:ices: Movie theaters offer rentals of their auditoriums for business rneetings and special events at off-peak

times, Fast-food chains o{I'er birthday party planning services to fill up slow demand periods

lf Your N.4ake a Reser\./atior-l, Sho\N LJp or PaYl Walt Disney World's better restaurants established a

policy whereby if you make a dinner reservation and do

not cancel at least 24 hours in advance, they will charge

your credit card S'l 0 per person. This no-show policy

applies to 19 Disney park restaurants'The most upscale

Disney restaurant charges 5ZS per person. Restaurant

(service) capacity is perishable, and Disney handles the

risks of idle service capacity by overbooking and/or

charging fees for abrupt cancellations and no-shows!6

The lWlysteries of Hotel RerzenLre l\./lanagernent How can the cost of the same hotel room, at the same property, vary dramatically from one day to the next? Historically, hotels have just looked at how many reservations they have or what competitors are chargirrg.

Today, they use analytics to decide how to price rooms based on the market they are in, special events or holidays that may affect demand, room differences, popularity

on different days of the week, and the guest mix at any point in time, and even weather forecasts. Like most hotei brands, Best Western revenue managers rely on analytical

methodology that includes forecasting demand, scanninl; websites like Expedia.com to see what their competitors

are charging, and comparing their own booking trends. For example, properties might pay more for a standard

room with a king-size bed on a weekdayfor solo business

travelers, while a room with two queen beds may be priced

higher on weekends when more families are traveling.T

reservation is a promise to provide o good or service at some future time and ploce. Typical examples

are reservations for hotel rooms, airline seats, and scheduled surgeries and operating rooms.

Reservations reduce the uncertainty for both the

good or service provider and the customer. With advrrnce knowledge of when customer demand will occur, operations managers can better plan their equipment and work{brce schedules and rely

less on fbrecasts.

10-4c Revenue Management Systems (RMSi

Many types o{'organizations manage perishable assets, suc r

as a hotel room, zln airline seat, a rental car, a sporting ever t or concert seat, a room on a cruise line, dre capacity of I

I

= o

between peak meal times. l- restaurant catering service or elec- Extendedhours also I DUrinp, the f irSt tric power generation, or brotrd- representaperipheral , I y"ra of RMS

cast advertising space. For such service; many supennarkets I remain open24l7 and implelnentation , service capacity, high utilization is encourage customers to 1. evenge sl at Nat iOna 1 the key to finirncjal success. slropduringlate-night /a^-^ D^*+.^1 :^^- ^^^, I Arevenuemanagement lrours to recluce a"r]r^ra Car Rent al increased I ty*"- (RMs) consistsof dynomic during peak tirnes. by $ S tt mil l ion . I methods to forecast demand, allocate

ments, decide when to overbook ond by how much, ond determine

whot price to charge different customer (price) c/osses. These four

components of RMS-forecasting, allocation, overbook- ing, and pricing-nrust work in unison if the objective is to rnaximize the rev- enue generated by a per- ishable asset. The ideas and methods surround- ing RMS are often called

peld management. Reve- nue management systems integrate a wide varie$' of decisions and data into a decision support system used mainly by service- providing businesses.

CHAPTER 10: Capacity Management 209

l- I rr you don't se1l or I t". the ticket, the I

revenue - generating opportunity for that seat cannot be recaPtured-it

is lost forever.

The earliest revenue management systems lbcusei

solely on overbooking-how many perishable asset s to sell in excess of physical capacity to optimally trade

off the cost of arr unsold asset versus the loss of goot -

will of having rror:e ar:rivals than assets, Modern RMS so{trvare simultaneously makes changes in the {brecast,

:rllocation, overbooking, and pricing decisions in a retr -

time operating system. Forecasts are corlstantly being

revised. Allocation involves segmenting the perishable

asset into target rnarket categories, such as first, busi-

ness, and coach classes in an airline {1ight' Each cla"s

is delined by its size (number of seats), price, advant e

purchase restrictions, and booking policies' Allocation

i, ^

r"nl-ti*e, ongoing rnethod that does not end unlil there is no more oPPor- tunity to maximize rev,r- nue (the night or concert

is over, the airpltrne tak,:s

o{1). As happens wit h forecasts, bookings ar d time move forward; tJ,e target market categori,:s

ar:e redefined; and pric';s change in an attemPt .o ma,rimize revenue.

Many org:rnizatio rs I-rave exploited RMS tecli-

nologr. Marriott imProv, xl

its revenues by {825-$.lS million by using RNIS methods. Royal Carib- bean Cruise Lines ob- tained a revenue increase

in excess of $20 million lor

one year.8 Duing the fl 'st

year o{' RMS implemenl a-

tion, revenues :rt Natiorral Car Rental increased by $56 million.e

Theater seats ProviJe another good examl,le

seat

210 PART THREE: Managing operations and Supply Chains

of rnanaging a perishable asset. Revenue management

systems are often used to manage theater, stadiurn, and concert seats. If you don't sell or use the ticket, the revenue-generating opportunity for that seat cannot be

recaptured-it is lost {brever.

THEORY OF CONSTRAINTS

rhe Theory of Constraints (TOC) is o set of principles that

focuses on increosing total process throughput by maximizing the

utilization of all bottleneck work activities ond workstations. The

TOC w:rs introduced in a fictional novel, The Goal, by

Dr. Eliyahu M. Goldratt.r0 The philosophy and princi- ples of the TOC are valuable in understanding demand

and capacity mtrnagement.

The tr:aditiorral OM definition of throughput is the average number: of goods or services cornpleted per

time period by tr process. The TOC views throughput

diflerently: throughput is the amount of money generoted

per time period through actual soles. For most business or-

ganizations the goal is to maximize throughput, thereby

mnximizing cash flow. Inherent in this definition is that it

rni*es little sense to make a good or service until it can be

sold, and that excess inventory is wasteful'

In tlre TOC, d Gonstraint is onything in on organization

thot limits it from moving toward or achieving its goal' Constraints

determine the throughput of a facility because they limit

production output to their own capacity' There are two

Lasic types of'constraints: physical and nonphysical'

n physical constraint is associoted with the copacity of

a resource Such as a mochine, employee, or workstotion' Plrysical

constraints result in process bottlenecks' A bottleneck (BN) work activity is one thot effectively |imits the capacity

of the entire process. At a bottleneck, the input exceeds the

capacity, restricting the total output that is capable of be-

ing procluced. A nonbottleneck (NBN) work activity

is one in which idle capacity exists.

I nonphysical Gonstrainl is environmentol or organi' zotionol, such as low product demond or an inefficient management

policy or procedure. Inllexible work rules, inadequate labor

shlls, and poor mirnagement are all fbrms o{'constraints'

Ilemoving nonphysical constraints is not always possible'

Because the number of constraints is typically small. the TOC fbcuses on identifying them; managing

BN and NBN work activities care{uily; linking them to

the market to ensure an appropriate procluct mix; :rnd

scheduling the NBN resources to enhance ihroughput'

These principles are summarized in Exhibit 10'7'

In general, the TOC has been success{ul in many com-

panies. As the TOC evolved, it has been applied not only to

Move jobs through nonbottleneck workstations as fast as possible until the job reaches the bottleneck workstation.

At nonbottleneck workstations, idle time is acceptable if there is no work to do, and therefore resource utilizations mav brr low.

Use smaller order (also caljed lot or transfer batches) sizes at nonbottleneck workstations to keep work flowing to the bottleneck resources and eventually to the marketplace to generate sales.

An hour lost at a nonbottleneck resource has no effect on toral process or factory output and incurs no real cost.

manufhcturing but to other areas such as distribution, rnarketing, and human resource manage- ment. Binney and Srnith, malcer of Crayola crayons, and Procter & Garnble both use tlie TOC in their distribution e{forts. Binney and Smith had high inventory lev- els yet poor customer service. By

For m<>st business organ.Lzations the

goal ir; to maximize througirput , thereby

maximizj-ng cash f1ow.

Only the bottleneck workstations are critical to achieving process and factory objectives and should be scheduled first.

An hour lost at a bottleneck resource is an hour lost for the entire process or factory output.

Work-in-process buffer inventory should be placed in front of bottlenecks to maximize resource utilization at the bottleneck. Use large order sizes at bottleneck workstations to minrmrze setup time and maximize resource utilization.

Bottleneck workstations should work at all times to maximize throughput and resource utilization so as to generate cash from sales and achieve the company's goal.

Procter & Gamble reported $600 rnillion in savings tl'rrough inven- tory reduction and elimination of capital improvement tJrrough the TOC. A governrnent organi- zation that produces publications of labor statistics for the state of' Pennsylvania used the TOC to better match work tasks to work-

using the TOC to better position its distribution inven o- ries, it was able to reduce inventories and improve serwir:e.

ers to reduce idle labor and overtime requirements, and to increase throughput, job stabllity, and profitabilirytl

Kreisler Manufacturing Corporation is a small, family-run company that makes metal components for airplanes. lts clients include Pratt & Whitney, General Electric, Rolls Royce, and Mitsubishi. After learning about the TOC, managers

identified several areas of the factory, including the Inte 'rnal Machine Shop and Supplier Deliveries, as bottlenecks, and began to focus on maximizing throughput at these bottlenecks. Setups were videotaped to see exactly what was happening. lt was discovered that 60 percent of the time it took to complete a setup involved the worker looking for materials and tools. To remove this constraint, Kreisler assembled all the necessary materials and tools for setup into a prepackage "kitl'thus cutting 60 percent off the setup time.

Kreisler also created a "visual factory" by installing red, yellow, and green lights on every machine. lf a workstation is being starved or production stops, the operator turrs on the red light. lf there is a potential crisis or a risk of starving the constraint workstation, the worker turns c n the yellow light. lf all is running smoothly, the green light is on. Giving the machine operator control over these signa|; instilled a sense of ownership in the process and caught the atlention and interest of everyone in the factory. In the early star;es of implementing the TOC, there were many red lights; torlay they are green. By applying the TOC, on-time deliveries incr,:ased to 97 percent from 65 percent, and 'l 5 percent of the factory's "hidden capacity"was revealed and freed up. In additic'n, WIP inventory was reduced by 20 percent and is expected to be reduced by another 50 percent.l2

E

g

E

=

CHAPTER I0: Capacity Management 21 1

DISCUSSION OUESTIONS

1. Provide and discuss some examples of economies anrl diseconomies of scale in a college environment.

Define useful capacity measures {br a(n)

a. brewery. b. airline. c. movie theater. d. pizzarestaurant. e. amusement park.

4.

How might a college or urriversity with growing enrollment use the capacity expansion strategies in Exhibit 10.6? Discuss the pros and cons ol'each o{' these.

Briefly describe a business you are I'amiliar with and explain how it might use each of the five ways to adjust its short-term capacify levels.

How would you apply the Theory of Constraints to a quick-service autornobile oil change service? Explain.

PROBLEMS AND ACTIVITIES

Note: An asterisk denotes problems for which a spreadsheet template found at OM Online may be used.

1. As the assistant manager of a restaurant, how many servers will you need given tlie following infbrmation for Saturday night's dinner menu?

o Demtrnd (dinners serwed) = 100 dinners per hour o Server target utilization : 85 Percent r Service rate per server : 16 dinnershour What does the service rate per server assulne?

Explain.

2. Researclr and write a shoft paper (two pages maxirnum) on organizations that have successfully

used the focused factory concePt.

3.*Medical Solutions Inc. has the {bilowing claims it must complete in the next week. The jobs are as

{bllows:

Setup (Changeover) Time per

Consolidated Work Order Work Order Work Orders Quantity (Minutes)

Processing Time pel Panel (Seconds)

Claim Type

Setup Process- Number of Time per ing Time Claims to ClaimTyPe Per Claim Process (Hours) (Hours)

Model XVT-5 9,500 90 2.05 Case panels Model UYT-3 7,500 75 1.78 Case

Model KLY-6 10,800 150 4.31 Case

What is the total workload (demand) in hours for this

work order rnix? Horv many machines will it take to do this work in 1, 2, or 3 days? How might this process be improved?

Identify one example of a resource witlr a very low

average utilization rate, and a second example with

a very high average utilization rate. Consider both

service and manufhcturing organizations Write a

short (one page typed) paper that describes these

situations and their capacity implications.

Hickory Manu{acturing Company {brecasts tlle {bllowing demand for a product (in thousands of units) over the next 5 years:

Forecast demand 60 79 81 84 84

Currently the manufacturer has seven machines that

operate on a two-shi{t (eight hours each) basis. Twenty

days per year are available for scheduled maintenance

of equipment with no process output. Assume there

are 250 workdays in a year. Each manulactured good takes 25 minutes to produce.

a. What is the capacity o{'the factory? b. At what capaeity levels (percentage of norrnal

capacity) would the {irm be operating over the

Cancer treatment

Spinal injury

Hip replacement

l8 't2

9

3

1

2

0.9

1.6

0.7

Given process claim capacity of 40 hours o{'work,

can the workload be completed this week? Explain'

If not, what short-term capacity solution would you recornmend? Show all computations.

4.*Abbott Manuf'acturing produces plastic cases for

solar photovoltaic panels and has decided to combine

orders frorn customers to increase work order size,

and thereby tnake one large production run per

model type. Plastic injection molding machines are

used to rnake these parts, and it is time consuming to clean the machines and set them up again betweerr

production runs. Each molding machine and operator

works one 9-hour shllt with a one-hour lunch break and one-half hour for operator breaks.

212 PARTTHREE: Managing Operations and Supply Chains

next 5 years based on the for:ec:rsted dernand? (Hint: Compute the ratio o{ demand to capacity {br.- eacl.r year.)

c. Does the lirm need to buy more machines? I{ so, how many? When? If not, justify.

7. The roller coaster at Treasure Island Amusement Park consists of 16 cars, each of which can car'ry up to thr,:e passengers. Accolding to a time study, each run take; 2.5 minutes, and the time to unload and load riders i; 3.5 minutes. What is the maximum effective capacif o{'tlre system in number of passengers per hour?

8. Worthington Hills grccery store has five regultrr checkout lines and one express line (12 iterns or less). Based on a sampling study, it takes ll rninutes on tht avertrge for tr custorner to go through the regular line and 4 minutes to go through the express line, The store is open frorn 9 a,rn. to 9 p.m. daily.

a. What is the store's maximum capacity (custorners processed per day)?

b. What is the store's capacity by day of the week if tlre five regulzu checkout lines operate according to the following schedule? (The express line is always open.)

Hours/Day Mon Tue Wed fhur Fri Sat Sun 9-l2a.m. 1 1 'l 1 3 5 2 '12-4p.m. 2 2 2 2 3 5 4 4-6p.m. 3* 3 3 3 5 3 2 6-9o.m. 4 4 4 4 5 3 1 *A "3" means three regular checkout lines are open on Monday from 4 to 6 p.m.

9. Given the lbllowing data lbr Alberts f'abricating production area:

Fixed costs {br one shift

Unit variable cost

Selling price

: $80,000 =$7 = $12

Number of machines 6 Nurnber of worhng days in year 340 Processing tirne per unit : 40 minutes a, What is the annual capacity with a single eight-

hour shift?

b. What is the capacity with two shifts? c' what is the break-even volume witb a sinsle-shi{i

operation?

d. What is the maximum revenue with a single shi{i? e. What is the break-even volume with a two-shift

ooeration?

10. For the Solved Problern in this chapter involving processing driver's licenses at the Archer County Coufthouse, if 40 drivers are to be pr:ocessed each hour, how many clerks and typists should be hired

assuming an 90 percent target utilization rate?

11. Due to county and state budget cuts, Archer County Courthouse (see the Solved Problern in this chapter) now has only two clerks and two typists. All other inforrnation relnains the same. What is the new labor utilization for each labor type, and where is the bottleneck in this tl-rree-stage process? What is the impact of your analysis on custorner service? How might the job and process design be irnproved?