6.7 Use of Safety Stock

When the EOQ assumptions are met, it is possible to schedule orders to arrive so that stockouts are completely avoided. However, if the demand or the lead time is uncertain, the exact demand during the lead time (which is the ROP in the EOQ situation) will not be known with certainty. Therefore, to prevent stockouts, it is necessary to carry additional inventory called safety stock.

When demand is unusually high during the lead time, you dip into the safety stock instead of encountering a stockout. Thus, the main purpose of safety stock is to avoid stockouts when the demand is higher than expected. Its use is shown in Figure 6.7. Note that although stockouts can often be avoided by using safety stock, there is still a chance that they may occur. The demand may be so high that all the safety stock is used up, and thus there is still a stockout.

One of the best ways to implement a safety stock policy is to adjust the reorder point. In the EOQ situation where the demand and lead time are constant, the reorder point is simply the amount of inventory that would be used during the lead time (i.e., the daily demand times the lead time in days). This is assumed to be known with certainty, so there is no need to place an order when the inventory position is more than this. However, when the daily demand or the lead time fluctuates and is uncertain, the exact amount of inventory that will be used during the lead time is uncertain. The average inventory usage during the lead time should be computed, and some safety stock should be added to this to avoid stockouts. The reorder point becomes

A pair of graphs showing the effect of having safety stock.

Figure 6.7 Use of Safety Stock

ROP=(Average demand during lead time) + (Safety stock)ROP=(Average demand during lead time) + SS
(6-15)

where

SS=safety stock

How to determine the correct amount of safety stock is the only remaining question. Two important factors in this decision are the stockout cost and the holding cost. The stockout cost usually involves lost sales and lost goodwill, which result in loss of future sales. If the holding cost is low but the stockout cost is high, a large amount of safety stock should be carried to avoid stockouts because it costs little to carry this and stockouts are expensive. On the other hand, if the stockout cost is low but the holding cost is high, a lower amount of safety stock would be preferred because having a stockout would cost very little but having too much safety stock would result in a much higher annual holding cost.

How is the optimum stock level determined? If demand fluctuates, the lead time is constant, and both the stockout cost per unit and the holding cost per unit are known, the use of a payoff/cost table might be considered. With only a small number of possible demand values during the lead time, a cost table could be constructed in which the different possible demand levels would be the states of nature and the different amounts of safety stock would be the alternatives. Using the techniques discussed in Chapter 3, the expected cost could be calculated for each safety stock level, and the minimum cost solution could be found.

However, a more general approach is to determine what service level is desired and then find the safety stock level that would accomplish this. A prudent manager will look at the holding cost and the stockout cost to help determine an appropriate service level. A service level indicates what percentage of the time customer demand is met. In other words, the service level is the percentage of time that stockouts are avoided. Thus,

Service level=1Probability of a stockout

or

Probability of a stockout=1Service level

Once the desired service level is established, the amount of safety stock to carry can be found using the probability distribution of demand during the lead time.

Safety Stock with the Normal Distribution

Equation 6-15 provides the general formula for determining the reorder point. When demand during the lead time is normally distributed, the reorder point becomes

ROP=(Average demand during lead time)+ZσdLT
(6-16)

where

Z=number of standard deviations for a given service levelσdLT=standard deviation of demand during the lead time

Thus, the amount of safety stock is simply ZσdLT. The following example looks at how to determine the appropriate safety stock level when demand during the lead time is normally distributed and the mean and standard deviation are known.

Hinsdale Company Example

The Hinsdale Company carries a variety of electronic inventory items, and these are typically identified by SKU. One particular item, SKU A3378, has a demand that is normally distributed during the lead time, with a mean of 350 units and a standard deviation of 10. Hinsdale wants to follow a policy that results in stockouts occurring only 5% of the time on any order. How much safety stock should be maintained, and what is the reorder point? Figure 6.8 helps visualize this example.

From the normal distribution table (Appendix A), we have Z=1.65:

ROP=(Average demand during lead time) + ZσdLT=350 + 1.65(10)=350 + 16.5 = 366.5 units (or about 367 units)

So the reorder point is 366.5, and the safety stock is 16.5 units.

Calculating Lead Time Demand and Standard Deviation

If the mean and standard deviation of demand during the lead time are not known, they must be calculated from historical demand and lead time data. Once these are found, Equation 6-16 can be used to find the safety stock and reorder point. Throughout this section, we assume that lead time is in days, although the same procedure can be applied to weeks, months, or any other time period. We will also assume that if demand fluctuates, the distribution of demand each day is identical to and independent of demand on other days. If both daily demand and lead time fluctuate, they are also assumed to be independent.

A bell curve that helps illustrate how much safety stock should be maintained based on demand.

Figure 6.8 Safety Stock and the Normal Distribution

There are three situations to consider. In each of the following ROP formulas, the average demand during the lead time is the first term and the safety stock (ZσdLT) is the second term.

  1. Demand is variable but lead time is constant:

    ROP=d¯L+Z(σdL)
    (6-17)

    where

    d¯=average daily demandσd=standard deviation of daily demandL=lead time in days
  2. Demand is constant, but lead time is variable:

    ROP=dL¯+Z(dσL)
    (6-18)

    where

    L¯=averagelead time σL=standard deviation of lead timed=dailydemand
  3. Both demand and lead time are variable:

    ROP=d¯L¯+Z(L¯σd2+d¯2σL2)
    (6-19)

Notice that the third situation is the most general case and the others can be derived from that. If either demand or lead time is constant, the standard deviation and variance for that would be 0, and the average would just equal the constant amount. Thus, the formula for ROP in situation 3 can be simplified to the ROP formula given for that situation.

Hinsdale Company Example, Continued

Hinsdale has decided to determine the safety stock and ROP for three other items: SKU F5402, SKU B7319, and SKU F9004.

For SKU F5402, the daily demand is normally distributed, with a mean of 15 units and a standard deviation of 3. Lead time is exactly 4 days. Hinsdale wants to maintain a 97% service level. What is the reorder point, and how much safety stock should be carried?

From Appendix A, for a 97% service level Z=1.88. Since demand is variable but lead time is constant, we find

ROP=d¯L+Z(σdL)=15(4)+1.88(34)=15(4)+1.88(6)=60+11.28=71.28

So the average demand during the lead time is 60, and the safety stock is 11.28 units.

For SKU B7319, the daily demand is constant at 25 units per day, and the lead time is normally distributed, with a mean of 6 days and a standard deviation of 3. Hinsdale wants to maintain a 98% service level on this particular product. What is the reorder point?

From Appendix A, for a 98% service level Z=2.05. Since demand is constant but lead time is variable, we find

ROP=d¯L+Z(dσL)=25(6)+2.05(25)(3)=150+2.05(75)=150+153.75=303.75

So the average demand during the lead time is 150, and the safety stock is 154.03 units.

For SKU F9004, the daily demand is normally distributed, with a mean of 20 units and a standard deviation of 4, and the lead time is normally distributed, with a mean of 5 days and a standard deviation of 2. Hinsdale wants to maintain a 94% service level on this particular product. What is the reorder point?

From Appendix A, for a 94% service level Z=1.55. Since both demand and lead time are variable, we find

ROP=d¯L¯+Z(L¯σd2+d¯2σL2)=(20)(5)+1.55(5(4)2+(20)2(2)2)=100+1.551680=100+1.55(40.99)=100+63.53=163.53

So the average demand during the lead time is 100, and the safety stock is 63.53 units.

As the service level increases, the safety stock increases at an increasing rate. Table 6.5 illustrates how the safety stock level would change in the Hinsdale Company (SKU A3378) example for changes in the service level. As the amount of safety stock increases, the annual holding cost increases as well.

Calculating Annual Holding Cost with Safety Stock

When the EOQ assumptions of constant demand and constant lead time are met, the average inventory is Q/2, and the annual holding cost is (Q/2)Ch. When safety stock is carried because demand fluctuates, the holding cost for this safety stock is added to the holding cost of the regular inventory to get the total annual holding cost. If demand during the lead time is normally distributed and safety stock is used, the average inventory on the order quantity (Q) is still Q/2, but the average amount of safety stock carried is simply the amount of the safety stock (SS) and not one-half this amount. Since demand during the lead time is normally distributed, there would be times when inventory usage during the lead time exceeded the expected amount and some safety stock would be used. But it is just as likely that the inventory usage during the lead time would be less than the expected amount and the order would arrive while some regular inventory remained in addition to all the safety stock. Thus, on the average, the company would always have this full amount of safety stock in inventory, and a holding cost would apply to all of this. From this, we have

Table 6.5 Safety Stock for SKU A3378 at Different Service Levels

SERVICE LEVEL (%) Z VALUE FROM NORMAL CURVE TABLE SAFETY STOCK (UNITS)
90 1.28 12.8
91 1.34 13.4
92 1.41 14.1
93 1.48 14.8
94 1.55 15.5
95 1.65 16.5
96 1.75 17.5
97 1.88 18.8
98 2.05 20.5
99 2.33 23.3
99.99 3.72 37.2

Total annual holding cost = Holding cost of regular inventory + Holding cost of safety stock

THC=Q2 Ch+(SS)Ch
(6-20)

where

THC=total annual holding costQ=order quantityCh=holding cost per unit per yearSS=safety stock

In the Hinsdale example for SKU A3378, let’s assume that the holding cost is $2 per unit per year. The amount of safety stock needed to achieve various service levels is shown in Table 6.5. The holding cost for the safety stock would be these amounts times $2 per unit. As illustrated in Figure 6.9, this holding cost would increase extremely rapidly once the service level reached 98%.

USING EXCEL QM FOR SAFETY STOCK PROBLEMS

To use Excel QM to determine the safety stock and reorder point, select Excel QM from the Add-Ins tab, and select Inventory—Reorder Point/Safety Stock (normal distribution). Enter a title when the input window appears, and click OK. Program 6.4A shows the input screen and formulas for the first three Hinsdale Company examples. Program 6.4B shows the solution.

A line graph illustrates the relationship between increasing service level and increasing inventory carrying costs.

Figure 6.9 Service Level Versus Annual Carrying Costs

Screenshot of Excel QM for three different scenarios for Hinsdale Safety Stock.

Program 6.4A Excel QM Formulas and Input Data for the Hinsdale Safety Stock Problem

Screenshot showing solutions to the three Hinsdale examples. Data tables are replicated from the previous figure, and results tables are now filled in with calculations.

Program 6.4B Excel QM Solutions for the Hinsdale Safety Stock Problem

..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset