A capacity cushion is the amount of capacity less than expected demand.

Capacity Planning

Chapter outline

12.1 Facilities decisions

12.2 Facilities strategy

12.3 Sales and operations planning definition

12.4 Cross-functional nature of S&OP

12.5 Planning options

12.6 Basic aggregate planning strategies

12.7 Aggregate planning costs

12.8 Aggregate planning example

12.9 Key points and terms

In this chapter, we discuss capacity decisions related to carrying out the produc- tion of goods and services. Firms make capacity planning decisions that are long range, medium range, and short range in nature. These decisions follow naturally from supply chain decisions that already have been made and forecasting infor- mation as an input.

Capacity decisions must be aligned with the operations strategy of a firm. The operations strategy provides a road map that is used in making supply chain deci- sions to create a network of organizations whose work and output are used to satisfy customers' product and service needs. Capacity decisions are based on forecasted estimates of future demand. For example, operations and marketing collaborate to develop a forecast for demand for resort spa services before the re- sort makes capacity planning decisions regarding the appropriate facility and staff sizes for the spa.

As was discussed in the previous chapter, long-range decisions are concerned with facilities and process selection, which typically extend about one or more years into the future. The first part of this chapter describes facilities decisions and a strategic approach to making them. In this chapter we also deal with medium- range aggregate planning, which extends from six months to a year or two into the future. The next chapter discusses short-range capacity decisions of less than six months regarding the scheduling of available resources to meet demand.

Facilities, aggregate planning, and scheduling form a hierarchy of capacity deci- sions about the planning of operations extending from long, to medium, to short range in nature. First, facility planning decisions are long term in nature, made to

285

286 Part Four Capacity and Scheduling

FIGURE 12.1 Hierarchy of capacity decisions.

0

Scheduling

.. 6 12

Months Planning Horizon

18 24

obtain physical capacity that must be planned, developed, and constructed bef011! its intended use. Then, aggregate planning determines the workforce level and p:ro-- duction output level for the medium term within the facility capacity available Finally, scheduling consists of short-term decisions that are constrained by aggregall! planning and allocates the available capacity by assigning it to specific activities.

This hierarchy of capacity decisions is shown in Figure 12.1. Notice that the deci- sions proceed from the top down and that there are feedback loops from the b~ tom up. Thus, scheduling decisions often indicate a need for revised aggrega~ planning, and aggregate planning also may uncover facility needs.

We define capacity (sometimes referred to as peak capacity) as the maximum output that can be produced over a specific period of time, such as a day, week, or year. Capacity can be measured in terms of output measures such as number ot units produced, tons produced, and number of customers served over a specified period. It also can be measured by physical asset availability, such as the number of hotel rooms available, or labor availability, for example, the labor available fO£ consulting or accounting services.

Estimating capacity depends on reasonable assumptions about facilities, equip- ment, and workforce availability for one, two, or three shifts as well as the operat- ing days per week or per year. If we assume two eight-hour shifts are available for five days per week all year, the capacity of a facility is 16 X 5 = 80 hours per week and 80 X 52 = 4160 hours per year. However, if the facility is staffed for only one shift, these capacity estimates must be halved. Facility capacity is not available un- less there is a workforce in place to operate it.

Utilization is the relationship between actual output and capacity and is de- fined by the following formula:

Actual output Utilization = C . X 100%

apaCity

The utilization of capacity is a useful measure for estimating how busy a facility is or the proportion of total capacity being used . It is almost never reasonable to plan for 100 percent utilization since spare (slack) capacity is needed for planned and unplanned events. Planned events may include required maintenance or equipment replacement, and unpla1med events could be a late delivery from a supplier or unexpected demand.

Chapter 12 Capacity Planning 287

-

Operations Leader Bridge Collapse Illustrates Need for Emergency Capacity

The I-35W Mississippi River bridge near downtown Mi nneapolis and the University of Minnesota col- lapsed during the evening rush hour on August 1, 2007. Seventy-five city, county, state, and federal agencies were involved in the rescue effort. Such cata- strophic events illustrate the necessity for capacity availability in emergency services.

The Minneapolis police and fire departments were among the first responders to the disaster, and while they generally have excess capacity, utilization of

those services was temporarily pushed to the maxi- mum. Suburban police and fire units, with their own available capacity, were brought in to handle other demands for those services that were unre- lated to the bridge collapse.

While a majority of injured victims were treated at Hennepin County Medical Center, nine other area hospitals also treated victims. Such collaboration among hospital emergency departments is necessary during large disasters as no single hospital has enough capacity to absorb these demand spikes.

The excellent working relationships among agencies that had developed through joint train- ing, planning, and previous emergency incidents were cited as one of the primary reasons that re- sponse and recovery operations went smoothly. As one rescue leader commented, "We didn't view it as a Minneapolis incident; it was a city/county/state incident."

Source: Adapted from information compiled from several sources, including "I-35W Mississippi River Bridge," www.wikipedia.org, 2009.

Utilization rates vary widely by industry and firm. Continuous flow processes may have utilization near 100 percent. Facilities with assembly-line processes may set planned utilization at 80 percent to allow for flexibility to meet unexpected demand. Batch and job shop processes generally have even lower utilization. Emergency services such as police, fire, and emergency medical care often have fairly low utilization, in part so that they can meet the demands placed on them during catastrophic events. The Operations Leader box tells the story of the Inter- state 35W bridge collapse in Minneapolis, Minnesota, as an example of the need for capacity from a variety of organizations during emergencies.

It is possible in the short term for a firm to operate above 100 percent utilization. Overtime or an increased work-flow rate can be used in the short term to meet highly variable or seasonal demand. Mail and package delivery services often use these means to increase work output before major holidays. However, firms cannot sustain this rapid rate of work for more than a short period. Worker burnout, de- layed equipment maintenance, and increased costs make it undesirable to operate at very high utilization over the medium or long term for most firms.

In addition to the theoretical peak capacity, there is an effective capacity that is obtained by subtracting downtime for maintenance, shift breaks, schedule changes, absenteeism, and other activities that decrease the capacity available. The effective capacity, then, is the amount of capacity that can be used in planning for actual facility output over a period of time. To estimate effective capacity for the previously described two-shift facility, we must subtract hours for planned and unforeseen events.

l li

I I

288 Part Four Capacity and Scheduling

12.1 FACILITIES DECISIONS

"OM: Featuring St. Alexius

Medical Center," Vol. X

Facilities decisions, the longest-term capacity planning decisions, are of gm1 importance to a firm. These decisions place physical constraints on the amo.- that can be produced, and they often require significant capital investm~ Therefore, facilities decisions involve all organizational functions and often made at the highest corporate level, including top management and the b~ of directors.

Firms must decide whether to expand existing facilities or build new ones. _ we discuss the facilities strategy below, we will see there are trade-offs that mm be considered. Expanding current facilities may provide location conveniences current employees but may not be the best location in the long-run. Alternativell new facilities can be located near a larger potential workforce but require duplial- tion of activities such as maintenance and training.

When construction is required, the lead time for many facilities decisions r~ from one to five years. The one-year time frame generally involves buildings - equipment that can be constructed quickly or leased. The five-year time frame volves large and complex facilities such as oil refineries, paper mills, steel millll and electricity generating plants.

In facilities decisions, there are five crucial questions:

1. How much capacity is needed? 2. How large should each facility be? 3. When is the capacity needed? 4. Where should the facilities be located?

5. What types of facilities/ capacity are needed?

The questions of how much, how large, when, where, and what type can be s~ rated conceptually but are often intertwined in practice. As a result, facilities deci- sions are exceedingly complex and difficult to analyze.

In the next section, these five types of facilities decisions are considered in de- tail. We stress the notion of a facilities strategy, cross-functional decision ma:king and the relationship of facilities strategy to business strategy.

12.2 FACILITIES STRATEGY

In Chapter 2 it was noted that a facilities strategy is one of the major parts of - operations strategy. Since major facilities decisions affect competitive success they need to be considered as part of the total operations strategy, not simply as a series of incremental capital-budgeting decisions. Supporting the operatioot strategy also applies to other major strategic decisions in operations regardind process design, the supply chain, and quality management, as we have already noted.

A facilities strategy considers the amount of capacity, the size of facilities, the timing of capacity changes, facilities locations, and the types of facilities needed for the long run. It must be coordinated with other functional areas due to the necessary investments (finance), market sizes that determine the amount ot capacity needed (marketing), workforce issues related to staffing new facilities (human resources), estimating costs in new facilities (accounting), and technology decisions regarding equipment investments (engineering). The facilities strategy

Amount of Capacity

FACILITIES STRATEGY. This Bacardi Rum factory supplies the entire North American market from a single modern automated distillery in Puerto Rico.

Chapter 12 Capacity Planning 289

needs to be considered in an integrated fashion with these functional areas and will be affected by the following factors:

1. Predicted demand. Formulating a facility strategy requires a forecast of de- mand; ·techniques for making these forecasts were considered in the previous chap- ter. Marketing often is involved in forecasting future demand.

2. Cost of facilities. Cost is driven by the amount of capacity added at one time, the timing, and the location of capacity. Accounting and finance are involved in estimating future costs and cash flows from facility strategies.

3. Likely behavior of competitors. An expected slow competitive response may lead the firm to add capacity to grab the market before competitors become strong. In contrast, an expected fast competitive response may cause the firm to be more cautious in expanding capacity.

4. Business strategy. The business strategy may dictate that a company put more emphasis on cost, service, or flexibility in facilities choices. For exam- ple, a business strategy to provide the best service can lead to facilities with some excess capacity or several market locations for fast service. Other busi- ness strategies can lead to cost minimization or attempts to maximize future flexibility.

5. International considerations. As markets and supply chains continue to be- come more global in nature, facilities often are located globally. This involves not merely chasing cheap labor but locating facilities for the best strategic ad- vantage, sometimes to access new markets or to obtain desired expertise in the workforce.

One part of a facilities strategy is the amount of capacity needed. This is deter- min~d both by forecasted demand and by a strategic decision by the firm about how much capacity to provide in relation to expected demand. This can best be described by the notion of a capacity cushion, which is defined as follows:

Capacity cushion = 100% - Utilization

The capacity cushion is the difference between the output that a firm could achieve and the real output that it produces to meet demand. Since capacity utilization reflects the output required to satisfy demand, a positive cushion means that there is more capacity available than is required to satisfy demand. Zero cushion means that the average demand equals the capacity available.

290 Part Four Capacity and Scheduling

Size of Facilities

The decision regarding a planned amount of cushion is strategic in na~ The capacity cushion should be planned into capacity decisions, as the cushiOIIC will affect service levels as well as the firm's ability to respond to unexpectel situations.

Three strategies can be adopted with respect to the amount of capacity cushi• 1. Large cushion. In this strategy, a large positive capacity cushion, with ca-

pacity in excess of average demand, is planned. The firm intentionally has more capacity than the average demand forecast. This type of strategy is ap- propriate when there is an expanding market or when the cost of buildinfl and operating capacity is inexpensive relative to the cost of running out capacity. Electric utilities adopt this approach, since blackouts and brown-- outs are generally not acceptable. Firms in growing markets may adopt a positive capacity cushion to enable them to capture market share ahead 01 their competitors. Also, a large cushion can help a firm meet unpredictabll customer demand, for example, for new technologies that very quickly be- come popular. Firms using a make-to-order process usually have a signifiJ cant capacity cushion.

2. Moderate cushion. In this strategy, the firm is more conservative with respect to capacity. Capacity is built to meet the average forecasted demand comfort- ably, with enough excess capacity to satisfy unexpected changes in demand as long as the changes are not hugely different from the forecast. This strategy is used when the cost (or consequences) of running out is approximately in bal- ance with the cost of excess capacity. •

3. Small cushion. In this strategy, a small or nearly zero capacity cushion is planned to maximize utilization. This strategy is appropriate when capacity is very expensive, relative to stockouts, as in the case of oil refineries, paper mills, and other capital-intensive industries. These facilities operate profitably only at very high utilization rates between 90 and 100 percent. While this strategy tends to maximize short-run earnings, it can be a disadvantage if competitors adopt larger capacity cushions. Competitors will be able to meet any demand in excess of a firm's capacity. Make-to-stock processes are likely to plan a small capacity cushion using this strategy.

When planning the capacity cushion, firms assess the probability of various levels of demand and then use those estimates to make decisions about planned increases or decreases in capacity. For example, suppose a firm has capacity to produce 1200 units, SO percent probability of 1000 units of demand, and SO percent probability of 800 units of demand. Then average demand is estimated to be (.S X 1000) + (.S X 800) = 900 units. Producing 900 units results in a (900/ 1200) X 100% = 7S% utilization rate. Based on existing capacity, the cushion is (100% - 7S%) = 2S%.

The solved problems section at the end of the chapter provides an example of how to compute the capacity cushion by using probabilities of demand, existing levels of capacity, and costs of building capacity. This method provides a quantita- tive basis for estimating the amount of capacity cushion that may be required.

After deciding on the amount of capacity to be provided, a facilities strategy must address how large each unit of capacity should be. This is a question involving economies of scale, based on the notion that large facilities are generally more economical because fixed costs can be spread over more units of production.

FIGURE 12.2 Optimum facility size.

Timing of Facility Decisions

Unit Cost

Economies of scale

Facility Size (units produced per year)

Chapter 12 Capacity Planning 291

Economies of scale occur for two reasons. First, the cost of building and operating large production equipment does not increase linearly with volume. A machine with twice the output rate gener- ally costs less than twice as much to buy and operate. Also, in larger facilities the overhead related to manag- ers and staff can be spread

over more units of production. As a result, the unit cost of production falls as facil- ity size increases when scale economies are present, as shown in Figure 12.2.

This is a good news-bad news story, for along with economies of scale come diseconomies of scale. As a facility gets larger, diseconomies can occur for several reasons. First, logistics diseconomies are present. For example, in a manufacturing firm, one large facility incurs more transportation costs to deliver goods to mar- kets than do two smaller facilities that are closer to their markets. In a service firm, a larger facility may require more movement of customers or materials, for exam- ple, moving patients around a large hospital or moving mail through a regional sorting center. Diseconomies of scale also occur because coordination costs in- crease in large facilities. As more layers of staff and management are added to manage large facilities, costs can increase faster than output. Furthermore, costs related to complexity and confusion rise as more products or services are added to a single facility. For these reasons, the curve in Figure 12.2 rises on the right-hand side due to diseconomies of scale.

As Figure 12.2 indicates, there is a minimum unit cost for a certain facility size. This optimal facility size will depend on how high the fixed costs are and how rapidly diseconomies of scale occur. As an example, Hewlett-Packard tends to op- erate small plants of fewer than 300 workers, with relatively low fixed costs. This plant size helps Hewlett-Packard encourage innovation in its many small product lines. By contrast, IBM operates very large plants of 5000 to 10,000 workers. IBM plants tend to be highly automated, and they use a decentralized management ap- proach to minimize diseconomies of scale. Each firm seems to have an optimal facility size, depending on its cost structure, product/ service mix, and particular operations strategy, which may emphasize cost, delivery flexibility, or service. Cost is, after all, not the only factor that affects facility size.

Another element of facilities strategy is the timing of capacity additions. There are basically two opposite strategies here.

1. Preempt the competition. In this strategy, the firm leads by building capacity in advance of the needs· of the market. This strategy provides a positive capacity cushion and may actually stimulate the market while at the same time prevent- ing competition from coming in for a while. Apple Inc. used this strategy in the early days of the personal computer market. Apple built capacity in advance of demand and had a lion's share of the market before competitors moved in. Apple still uses this strategy today in building massive capacity and invento- ries in advance of new product launches for the iPad and iPhone.

292 Part Four Capacity and Scheduling

Facility Location

"Wind Farm Capacity," Vol. XVII

Types of Facilities

2. Wait and see. In this strategy, the firm waits to add capacity until demand develops and the need for more capacity is clear. As a result, the company lags market demand, using a lower-

·risk strategy. A small or negative capaci cushion can develop, and a loss of poten · market share may result. However, this stra~ egy can be effective because superior market- ing channels or technology can allow the follower to capture market share. For examplE;. in the 1980s, IBM followed the leader (Apple in the personal computer market but was able to take away market share because of its supe- rior brand image, size, and market presence. In

contrast, U.S. automobile companies followed the wait-and-see strategy, to their chagrin, for small autos. While U.S. automakers waited to see how demand for small cars would develop, Japanese manufacturers grabbed a dom- inant position in the U.S. small-car market.

Facility location decisions have become more complex as globalization has ex- panded the options for locating capacity and developing new markets. For example. Starbucks may choose to build facilities in regions with heavy coffee drinkers, com- peting for customers there. Alternatively, it may choose to locate facilities in regions where people typically do not consume coffee and attempt to create demand foc their product and service. Starbucks's U.S. competitor Caribou Coffee goes to great lengths to find facilities locations on the right-hand side of the road of morning traf- fic, because customers are more likely to stop frequently when they can pull over on the right but are less willing to make cumbersome left-hand turns in heavy traffic!

Location decisions are made by considering both quantitative and qualitative fac- tors. Quantitative factors that affect the location decision may include return on in- vestment, net present value, transportation costs, taxes, and lead times for delivering goods and services. Qualitative factors can include language and norms, attitudes among workers and customers, and proximity to customers, suppliers, and competi- tors. Front office services in particular must often locate near customers for the customers' convenience, and so this factor may trump most others in deciding where to locate new facilities. Examples include banks, grocery stores, and restaurants.

Firms often compare potential locations by weighting the importance of each factor that is relevant to the decision and then scoring each potential location on those factors. Then, multiplying the factor weight by the location score allows cal- culation of a weighted-average score for each site. This score provides insight on how well each potential site meets the needs of the firm and may be used to make the final facility location decision.

The final element in facility strategy considers the question of what the firm plans to accomplish in each facility. There are four different types of facilities:

1. Product-focused (55 percent)

2. Market-focused (30 p ercent) 3. Process-focused (10 percent) 4. General-purpose (5 p ercent)

Chapter 12 Capacity Planning 293

The figures in parentheses indicate the approximate percentages of companies in the Fortune 500 using each type of facility.

Product-focused facilities produce one family or type of product or service, usually for a large market. An example is the Andersen Window plant, which pro- duces various types of windows for the entire United States from a single product- focused plant. Product-focused plants often are used when transportation costs are low or economies of scale are high. This tends to centralize facilities into one location or a few locations. Other examples of product-focused facilities are large bank credit card processing operations and auto leasing companies that process leases for cars throughout the United States from a single site.

Market-focused facilities are located in the markets they serve. Many service facilities fall into this category since services generally cannot be transported. Plants that require quick customer response or customized products or that have high transportation costs tend to be market focused. For example, due to the bulky nature and high shipping costs of mattresses, most production plants are located in regional markets. International facilities also tend to be market focused because of tariffs, trade barriers, and potential currency fluctuations.

Process-focused facilities have one technology or at most two. These facilities frequently produce components or subassemblies that are supplied to other facili- ties for further processing. This is common in the auto industry, in which en- gine plants and transmission plants feed the final assembly plants. Process-focused plants such as oil refineries can make a wide variety of products within the given process technology.

Operations Leader Strategic Capacity Planning at BMW '

Mathematical optimization models (e.g., mixed linear programming) can be helpful in evaluating various ca- pacity strategies. Given a forecast of demand over the next several years, the models will determine the min- imum cost plan for meeting the demand.

An example from BMW helps explain this process. BMW used a 12-year future planning horizon to rep- resent the typical development and production life

cycle for new BMW designs. The mathematical model calculated the supply of finished products that should be produced by each plant to meet the forecasts in all global markets.

As a result, the model determined how much of each product should be produced in each plant over the next 12 years. All possible locations, amounts, and types of BMWs were considered by the mathe- matical model in arriving at the minimum cost plan.

This type of analysis is very helpful in setting overall capacity strategies for how much, how large, where, when, and what type. For example, the resulting strat- egy might be to use distributed production to pro- duce in each national market the amount sold there or a more centralized production and export strategy. Due to the many assumptions made, the models do not determine the final capacity strategy; however, they are very helpful in evaluating many different possibilities.

Source: Fleischmann B., Ferber S. and Henrich P., "Strategic planning of BMW's Global Production Network," Interfaces, 2006, 36(3): 194-208 ..

li II

II II 11

II

II

294 Part Four Capacity arul Scheduling

General-purpose facilities may produce several types of products and sel'- vices, using several different processes. For example, general-purpose facilities a.rr used to manufacture furniture and to provide consumer banking and investmenll services. General-purpose facilities usually offer a great deal of flexibility in terms of the mix of products or services that are produced there. They sometimes are used by firms that do not have sufficient volume to justify more than one facili~ Larger firms often focus their faCilities according to product, market, or procesa using the focused factory approach described in Chapter 4.

We have shown how the facilities strategy can be developed by considerind questions of capacity, size of facilities, timing, location, and types of facilities. Mathematical optimization models can often be helpful in answering these five strategic questions as shown in the BMW Operations Leader box. We now shift from long-term facilities decisions to medium-term decisions regarding how capacity, once built, is used.

12.3 SALES AND OPERATIONS PLANNING DEFINITION

Sales and operations planning (S&OP) is a term used by many firms to describe the aggregate planning process. Aggregate planning is the activity of matching supply of output with demand over the medium time range. The time frame is between six months and two years into the future, or an average of about one year. The term ag- gregate implies that the planning is done for a single overall measure of output or at most a few aggregated product categories. The aim of S&OP is to set overall output levels in the medium-term future in the face of fluctuating or uncertain demand.

We use a broad definition of S&OP with the following characteristics:

1. A time horizon of about 12 months, with updating of the plan on a periodic basis, perhaps monthly.

2. An aggregate level of demand for one or a few categories of product. The de- mand is assumed to be fluctuating, unc~rtain, or seasonal.

3. The possibility of changing both supply and demand variables. 4. A variety of management objectives, which might include low inventories,

good labor relations, low costs, flexibility to increase future output levels, and good customer service.