Wednesday, July 29, 2009

Demand Planning - DRP (Distribution Requirement Plan) - Part - IV

In this session, we will explore how DRP module works for Central Distribution Centre.

As stated earlier CDC act as a serving point to three Regional DCs. Hence, the Order Planned quantity (not Forecast Quantity or Net Requirement) of three regional DCs are aggregated to the CDC forecast quantity. For CDC we have assumed that the safety stock is nil. The Lot quantity for CDC is 300 units and lead time to get the stocks from CDC is 2 days.

We give below the DRP model used to compute the Order Planned quantity for CDC.

The Forecast Qty split is the aggregation of Order Planned Quantity of all three Regional DCs. Since the Safety stock is 0, Gross requirement is the same as Forecast Qty Split. Rest of the calculations are same, as explained in our earlier session.


- DRP is a scheduling and Stocking Algorithm. It replaces the forecasting mechanism above the base Inventory level.
- DRP does not determine the Lot Size and Safety stock. However Lot Size and Safety stocks are used as input to the DRP process
- DRP system can deal with market uncertainty through Safety stock and Lead time.

From the example, one can presume that DRP system is not that complicated as perceived. So far we are working under the assumption that there is no change in the demand (i.e., Forecast Quantity Split).

But in reality Market is more dynamic and hence the demand. What Happens when the actual demand differ from the forecast ? In such case should we follow the same DRP model as we worked out earlier. The DRP network in change in demand scenario is given below.

In the above scenario, the Regional DC 1 is facing shortage of stocks due to actual demand is more than expected demand (forecast). Regional DC 2 is facing excess stocks due to actual demand is less than expected demand. In such cases we have two options.

Option 1 – Manufacturing Plant should produce more to meet the Regional DC 1 requirement. Change in production schedule involves more cost and it is not viable option. Also the question arise what we are going to do with excess stock lying at Regional DC 2 ?

Option 2 – Is it possible to shift the stocks between the two Regional DC 1 & DC 2 so that excess stocks can be transferred to shortage DCs. This lead to re-deployment of existing inventory.
How do I re-deploy the inventory to take the maximum advantage of what inventory do we have ? The given below model explain about the re-deployment of inventory among the Regional DCs.

From the above diagram we can understand that the excess stocks from Regional DC 2 can be shifted to Local DC1 and DC2. Secondly even if we require more stocks then advise manufacturing plant to dispatch stocks from the existing safety stocks to Local DC1 & DC2 directly to avoid transit delay. Still we require more stocks then we can advise plant to produce more to meet the market requirements provided the cost is justifiable. If the production cost due to change in production schedule out weight the benefit from sales realization then it is not advisable to produce more items.

The Demand Planner is expected to take decisions on these issue in DRP system to meet the market demand. However he should be fully conversant with process mapping and IT architecture also. For example as per process if Regional DC 2 is expected to cater only Local DC3 & DC4 then Demand planner cannot transfer the stocks directly from Regional DC1 to Local DC1 & DC2. He has to transfer the stocks from Regional DC2 to Regional DC1. Regional DC1 will serve the Local DC1 and DC2 as per architecture.

So far we have discussed about Sales function modules like forecast and demand planning. In the next session we will focus our attention to Production function related topics like MRP (Material Requirement Planning) and MPS (Master Production Schedule), PPC (Production Planning & Control) etc.


  1. Shail, Thanks for your encouraging comment.

  2. Had a very gud undrstndg on DRP...very useful blog....request to keep it going.....

  3. How does centralized distribution centres reduce forecast error and the bullwhip effect?