Page 205 - Orlicky's Material Requirements Planning

184 PART 3 Managing with the MRP System

The average actual lead time of successive orders for a given item should, with ade-
quate capacity planning, approximate planned lead time. The average actual lead time of
all orders that are in process simultaneously is a function of capacity and the level of
work-in-process. Care must be taken when determining this lead time for purposes of
projecting work-in-process levels or of expressing the relationship between capacity,
work-in-process, and lead time algebraically because average actual lead time, when sim-
ply measured historically, will tend to be distorted. It will be inflated to the extent that it
includes orders with priorities that have dropped significantly subsequent to release, that
is, orders for items that had requirements deferred into the far future or for which
requirements disappeared entirely. More will be covered on this later.
The point of this discussion is that the old concept of a “good” or “accurate” lead
time, that is to say, an accurate planned lead time, must be discarded. Planned lead times
need not and should not necessarily equal actual lead times. Actual lead time is flexible.

SAFETY STOCK IN A NEW LIGHT
The venerable concept of safety stock needs to be rethought in light of the fact that actu-
al lead time is flexible and that modern MRP and time-phased order-point inventory
management systems have the ability to realign open-order due dates with shifting dates
of need. The traditional approach to inventory control assumes that lead time is fixed and
known and that only demand is variable. But what happens to this approach when it
turns out that actual lead time is, in fact, flexible when it can be made to expand and con-
tract with demand? What is to become of safety stock, and how is it to be calculated
1
now? When there is a system capable of replanning priorities by revising open-order due
dates, there certainly is no longer any need for safety stock to cover the period of planned
replenishment lead time.
This would seem to suggest that the traditional techniques of determining safety
stock are obsolete and that new techniques and a new approach are needed. Safety stock,
where it applies, should be susceptible to reduction, across the board, without an adverse
2
effect on service. Safety stock on the item level is not, of course, normally planned by an
MRP system and is not intended to be. It is, however, planned under a time-phased order
point. In either case, the MRP logic common to both systems tends to defeat the purpose
of safety stock by preventing it from ever actually being used—if the system can help it.
This will be demonstrated in an example illustrated by Figures 10-1, 10-2, and 10-3.
The item in this example has a planned lead time of four periods, demand during
lead time is forecast as 40 units, and safety stock is 20. The order point therefore is 60.
Figure 10-1 shows, in graphic form, the inventory projection and the position of the
replenishment order at the time order point is reached. Depending on whether the item
is under statistical order point or time-phased order point (equivalent to MRP), the pro-

1 O. W. Wight, Oliver Wight Newsletter 12, May 1972.
2 A. O. Putnam, E. R. Barlow, and G. N. Stilian, Unified Operations Management. New York: McGraw-Hill, 1963, p.
183.