How MRP Calculates Requirements: A Step-by-Step Example

Most explanations of MRP stop at the definition: material requirements planning determines what to buy and make, and when.

That is true, and not very useful if you are trying to understand why MRP just told you to release a purchase order today for material you will not use for three weeks, or why it did not suggest anything for an item you are about to run out of.

The best way to understand MRP is to watch it calculate.

This guide walks through a complete MRP run for a simple product with a two-level bill of materials, one period at a time: gross requirements, scheduled receipts, projected on-hand, net requirements, lot sizing, safety stock, and lead-time offsetting. Then it covers the parts that cause trouble in real plants: exception messages, pegging, time fences, and the data problems that make MRP output wrong.

If you are looking for how MRP fits within ERP, start with our guide to ERP vs. MRP. This article is about the calculation itself.

The Three Inputs to MRP

Every MRP calculation depends on three sets of data:

• Demand: what needs to be shipped or built, and when. This usually comes from a master production schedule (MPS), customer orders, and forecasts.
• Bills of material: which components go into each item, and how many. See our guide to BOM management.
• Inventory status: what is on hand, what is already on order (open purchase orders and work orders), and planning parameters such as lead times, lot-size rules, and safety stock for each item.

MRP combines these to answer one question for every item, in every period: will we have enough, and if not, when must we order more?

The MRP Record: Key Terms

MRP calculates a time-phased record for each item, usually in weekly or daily buckets:

• Gross requirements: total demand for the item in the period, from customer orders or forecasts for finished goods, or from parent items’ planned orders for components.
• Scheduled receipts: open orders already released, such as purchase orders and work orders, and when they are due.
• Projected on-hand: expected inventory at the end of each period.
• Net requirements: the shortfall that must be covered by a new order.
• Planned order receipts: new orders MRP recommends, in the period they must arrive.
• Planned order releases: the same orders, moved earlier by the item’s lead time, which is when they must be placed or started.

The calculation, period by period:

Projected on-hand = Previous on-hand + Scheduled receipts + Planned order receipts − Gross requirements

When projected on-hand would fall below zero, or below safety stock, MRP creates a planned order to cover the gap.

The Example: A Table With a Two-Level BOM

A manufacturer builds a simple table. The bill of materials:

• Table (finished good, assembled in-house)
  • 1 × Top (purchased)
  • 4 × Leg (manufactured in-house)
    • 1 × Steel tube blank per leg (purchased raw material)

Planning data:

Demand for tables over the next six weeks: 50 in week 3, 80 in week 5, and 40 in week 6. It is currently week 1.

MRP works top-down, level by level, because a component’s requirements depend on its parent’s planned orders.

Step 1: The Finished Product (Table)

In week 3, demand of 50 exceeds the 20 on hand, so MRP needs 30 more. With a lot-for-lot rule, it plans exactly 30 to arrive in week 3. With a one-week lead time, the work order must be released in week 2. Weeks 5 and 6 work the same way.

The planned order releases, 30 in week 2, 80 in week 4, and 40 in week 5, become the demand for the table’s components.

Step 2: A Manufactured Component (Leg)

Each table needs 4 legs, so the leg’s gross requirements are 4 × the table’s planned order releases: 120 in week 2, 320 in week 4, and 160 in week 5. The leg also has an open work order for 100, due in week 1.

The open work order brings on-hand to 160 in week 1, which covers week 2’s 120 and leaves 40. In week 4, the requirement of 320 leaves a shortfall of 280, so MRP plans a work order for 280 legs, released in week 3. Week 5 needs another 160, released in week 4.

Notice that the table’s demand in week 5 became a leg requirement in week 4, which will become a steel requirement in week 3. Each level moves requirements earlier by the parent’s lead time. This is called lead-time offsetting, and it is why a customer order for week 5 can require action on raw material today.

Step 3: A Purchased Raw Material With a Lot-Size Rule (Steel Tube Blank)

One blank per leg, so the blank’s gross requirements equal the leg’s planned order releases: 280 in week 3 and 160 in week 4. The supplier sells blanks only in multiples of 250, and the lead time is 3 weeks.

The 300 on hand covers week 3, leaving 20. Week 4 needs 160, a shortfall of 140. Because the supplier sells in multiples of 250, MRP rounds up to 250, leaving 110 in inventory afterward.

With a 3-week lead time, the purchase order must be placed in week 1, which is now. This is the action a buyer sees on the MRP report: place a purchase order for 250 blanks this week.

Step 4: A Purchased Component With Safety Stock (Top)

One top per table, so the gross requirements equal the table’s planned order releases: 30 in week 2, 80 in week 4, and 40 in week 5. Tops have a 2-week lead time and a safety stock of 10, which means MRP plans to keep at least 10 on hand at the end of every period.

In week 4, 20 on hand minus 80 required would leave −60. To end the week with the 10-unit safety stock, MRP plans 70, not 60. In week 5, it plans 40 to stay at 10.

Safety stock is a buffer against variability in demand or supply, but it is not free. It permanently ties up inventory, and in MRP it makes every planned order slightly larger.

What the MRP Run Produced

From three customer demand quantities, MRP produced a complete, time-phased plan:

Only one order needs action today: the steel. Every other order can wait until its release week. MRP will recalculate before then, and if demand or inventory changes, the plan will change with it.

A real plant runs this calculation across thousands of items and many BOM levels, every time demand, inventory, or an open order changes. That is why MRP is software, not a spreadsheet.

Lot-Sizing Rules

The example used two rules. These are the most common:

• Lot-for-lot: order exactly the net requirement. Minimizes inventory, but can create many small orders.
• Fixed order quantity: always order the same amount, such as a full pallet or a standard batch size.
• Minimum, maximum, and multiple: respect a supplier’s minimum order, a storage or batch maximum, and pack-size multiples, like the 250-blank rule above.
• Period order quantity: combine several periods of requirements into one order, such as two weeks’ worth at a time.
• Economic order quantity (EOQ): balance ordering or setup cost against carrying cost to find a cost-minimizing quantity.

Lot sizes at one level ripple down. A table batch rounded up to a full run creates larger leg requirements, which create larger steel requirements.

Exception Messages

Planners rarely read every MRP record. They work from exception messages, which tell them where the existing plan no longer matches reality:

• Release: a planned order has reached its release date. In the example, the steel purchase order.
• Expedite: an open order needs to arrive earlier than it is currently due.
• De-expedite (push out): an open order is due earlier than it is needed.
• Cancel: an open order is no longer needed at all.
• Past due: a planned order should already have been released. If the steel supplier’s real lead time were 4 weeks instead of 3, the release date would fall before today, and MRP would flag it as past due.

A healthy plan has few exceptions. A plan with hundreds of them every day usually has a data problem, not a planning problem.

Pegging

Pegging traces a requirement back to its source. In the example, the 280-leg work order in week 4 pegs to the 80-table work order in week 4, which pegs to customer demand for 80 tables in week 5. (The 280 is net of the 40 legs left in stock, so most of that table order’s 320 legs are coming from this work order.)

Pegging answers the question a planner needs before expediting or cancelling anything: which customer order is this for, and what happens to it if this order is late?

Time Fences and Firm Planned Orders

MRP recalculates constantly, and a plan that changes every hour is impossible to execute. Two tools add stability:

• Time fences: inside a defined window, often the next few days or weeks, MRP does not automatically change the plan. Changes in that window require a planner’s decision.
• Firm planned orders: a planner can lock a planned order’s quantity and date, so MRP plans around it instead of recalculating it.

Why MRP Output Goes Wrong

When planners stop trusting MRP, the cause is almost always input data:

• Inaccurate inventory. If the system says 300 blanks and the shelf has 180, every downstream recommendation is wrong.
• Inaccurate BOMs. A missing or wrong quantity means MRP never plans for it. See our guide to BOM management.
• Unrealistic lead times. Lead times copied from a vendor’s catalog, or never updated, put every release on the wrong date.
• Stale open orders. Purchase orders and work orders that will never arrive, or arrived and were never received, make MRP think supply exists.
• Unrecorded scrap and consumption. Material used on the floor without a transaction disappears from reality but not from the system. See our guide to WIP inventory.
• Infrequent runs. A weekly or nightly MRP run is working from data that may be days old by the time anyone acts on it.

What MRP Does Not Do: Capacity

Classic MRP assumes infinite capacity. In the example, it planned a 280-leg work order and a 160-leg work order in consecutive weeks without asking whether the leg cell can actually produce 280 legs in a week.

That check belongs to capacity planning and scheduling. MRP decides what is needed and when; capacity planning decides whether the plant can do it; scheduling decides exactly when each operation runs. See our guides to capacity planning and production scheduling.

MRP in a Modern ERP

The calculation has not changed much since MRP was developed. What has changed is the data it runs on.

When MRP is part of the same system as sales orders, purchasing, inventory, and the shop floor, its inputs are current: demand reflects the order entered an hour ago, inventory reflects the last scan, and open orders reflect the last receipt. The plan is only as current as its data, so a system where those transactions happen in one place produces a more trustworthy plan than one that imports them overnight.

How Bizowie handles it: Bizowie Cloud ERP includes MRP that is always up to date, so purchasing and production plans reflect current demand, inventory, and open orders. Work orders can be created automatically or manually from sales activity and planned demand, including dependent work orders for multi-level subassemblies, and MRP reporting shows purchasing requirements and shortages. Because Bizowie also schedules production based on labor and machine capacity, with finite, infinite, and mixed capacity methodologies, the material plan and the capacity plan come from the same data.

For more, see our guide to ERP vs. MRP and our manufacturing ERP software buyer’s guide.

Frequently Asked Questions

What are the steps of an MRP calculation?

For each item, starting with finished goods and working down the BOM: determine gross requirements, add scheduled receipts, calculate projected on-hand, identify net requirements where on-hand falls below zero or safety stock, apply the lot-size rule to create planned order receipts, and offset them by lead time to get planned order releases. The parent’s releases become the components’ gross requirements.

What is the difference between gross and net requirements?

Gross requirements are total demand for an item in a period. Net requirements are what remains after subtracting available inventory and scheduled receipts, and accounting for safety stock. MRP creates planned orders only for net requirements.

What is lead-time offsetting?

Lead-time offsetting moves a planned order’s release date earlier than its due date by the item’s lead time. It is why demand for a finished product can require purchasing raw materials weeks earlier.

What is the difference between a planned order and a scheduled receipt?

A planned order is a recommendation that MRP can change on the next run. A scheduled receipt is an order already released, such as an open purchase order or work order, that MRP treats as committed supply.

How often should MRP be run?

Traditionally MRP was run weekly or nightly because of computing limits. Modern systems can recalculate much more often, or continuously, so plans reflect current orders and inventory. What matters most is that planners act on current data.

Does MRP consider capacity?

Classic MRP does not; it assumes infinite capacity. Capacity requirements planning and finite scheduling are used alongside MRP to check whether the plant can actually produce the planned orders.

Run It on Your Own Numbers

The example above has four items. Your plant has thousands. Bring us a real case:

• An item MRP keeps recommending that nobody believes.
• A multi-level product where a demand change ripples down the BOM.
• A supplier with minimums and pack sizes.
• A shortage that nobody saw coming.

We’ll show you how Bizowie would plan it, and if the platform is not a good fit for your operation, we’ll tell you that too.

See if Bizowie is right for your manufacturing operation.