Material requirements planning
Material Requirements Planning (MRP): inputs (MPS, BOM, inventory records), time‑phased netting, lot sizing, lead‑time offsetting, explosion to lower levels, with safety stock/lead time/rounding effects.
Drafted with Aria, reviewed by the AiCanCode.org team · Updated . How we write and check our notes. Spotted an error? Use Give Feedback at the bottom of the page.
Why it matters
MRP ensures the right materials are available at the right time and quantity so production stays on schedule with minimal inventory. It links demand for finished goods to time‑phased orders for components and purchases, cutting stockouts, expediting, and carrying cost.
Key ideas
- Material Requirements Planning (MRP): A time‑phased planning system that converts a Master Production Schedule (MPS) for finished items into planned order releases for components and raw materials using the Bill of Materials (BOM), inventory records, and lead times.
- Inputs:
- Master Production Schedule (MPS): Time‑bucketed planned production or shipment quantities for end items (or key subassemblies). Assumes the MPS is feasible and demand is known or decided by Production Planning and Control/Forecasting.
- Bill of Materials (BOM): Product structure listing parent–child usage quantities, alternates, and by‑products. Assumes exact usage per good unit, 100% yield unless scrap/yield adjustments are specified.
- Inventory records: On‑hand balance, scheduled receipts (open purchase/work orders due in a bucket), lead time, lot‑sizing rule, and safety stock or safety lead time per item.
- Outputs:
- Planned order receipts (PORcpt): Quantities to be received in a future bucket to meet net requirements.
- Planned order releases (PORrel): The same quantities offset earlier by lead time to trigger start/placement.
- Exception/action messages: expedite, de‑expedite, cancel, or reschedule recommendations; signals of shortages and excess.
- Time phasing and buckets:
- MRP is discrete‑time by buckets (often weeks). We treat time in weeks in examples (industry standard). SI base time is the second; convert if required for dimensional checks.
- Low‑level code (LLC): The lowest level in any BOM at which a part appears. MRP processes items in ascending LLC so that child requirements see the parent’s planned releases in the same run.
- Gross vs net requirements:
- Gross requirements (GR): Total demand in a bucket from the MPS (for end items) or from parents’ planned releases multiplied by usage (for components), without regard to inventory.
- Projected available balance (PAB): End‑of‑bucket on‑hand projection after planned receipts and issues.
- Net requirements (NR): Additional quantity needed in a bucket to ensure PAB does not fall below safety stock.
- Safety stock vs safety lead time:
- Safety stock (SS): A minimum PAB target each bucket to buffer variability in demand/lead time; implemented by increasing requirements by SS for netting.
- Safety lead time (SLT): Bringing orders earlier by SLT buckets instead of holding extra stock; reduces inventory but advances releases.
- Lead‑time offsetting: Planned releases are placed L buckets before planned receipts to respect item lead time L (routing/queue/purchase time assumed deterministic in basic MRP).
- BOM explosion:
- For a parent’s planned release in bucket t, a child’s gross requirement in the same bucket t equals parent release × usage (adjusted for yield/scrap if applicable). This propagates from top level (end items) down to raw materials.
- Scrap/yield factors:
- If an operation has scrap rate s (fraction lost), to obtain GR good units you must plan receipts GR/(1−s) before rounding/lot sizing. For downstream components, use the planned releases of the scrapped parent as the driver.
- Lot‑sizing rules:
- Lot‑for‑Lot (LFL): Order exactly the net requirement of the bucket. Minimizes average inventory; can cause many setups.
- Fixed Order Quantity (FOQ): Order in fixed multiples Q. Planned receipt is the smallest multiple of Q ≥ NR.
- Economic Order Quantity (EOQ): Classical cost‑based lot size Q* = sqrt(2DS/H) under steady demand and no time phasing; adapted in MRP via FOQ≈EOQ or Period Order Quantity.
- Period Order Quantity (POQ): Order to cover a fixed number of periods equal to EOQ divided by average demand per period.
- Heuristics for nonstationary demand: Silver–Meal (choose horizon T with minimum average cost per period), Part‑Period Balancing (PPB) (equate setup cost to holding of part‑periods). Exams may ask to compute 2–4 periods by these rules.
- MRP run types:
- Regenerative: Rebuild entire plan item‑by‑item periodically.
- Net change: Update only items affected by recent changes.
- Pegging: Ability to trace a child requirement back to the specific parent and MPS source.
- Assumptions in basic MRP:
- Deterministic lead times, known BOM and usage, infinite capacity (capacity checks handled separately by CRP), fixed yield unless specified, discrete review at bucket boundaries, accurate data, and no constraints on storage or minimum order amounts beyond lot‑sizing rule.
- Why real systems differ from ideal:
- Lead times vary with load; yields are stochastic; suppliers have MOQs and calendars; engineering changes modify BOMs; capacity and time fences limit changes; data inaccuracies cause nervousness. Hence MRP is often combined with safety stocks/lead times, firm planned orders, and CRP/finite scheduling for realism.
Formulas
PAB_t = PAB_{t−1} + SR_t + PORcpt_t − GR_t
- PAB_t: projected available balance at end of bucket t [piece]; SR_t: scheduled receipts due in t [piece]; PORcpt_t: planned order receipts in t [piece]; GR_t: gross requirements in t [piece]. Applies per item, per bucket, when PAB is tracked end‑of‑bucket; receipts positive, requirements negative.
NR_t = max(0, GR_t + SS − (PAB_{t−1} + SR_t))
- NR_t: net requirements in bucket t [piece]; SS: safety stock target [piece]. Applies with safety stock as a minimum end‑of‑bucket level and LFL logic before lot‑sizing.
PORcpt_t = ceil(NR_t / Q) × Q
- For FOQ lot sizing; Q: fixed lot size [piece]. If LFL, take Q → 1 (i.e., PORcpt_t = NR_t). For minimum order quantity M with increments I, adapt as PORcpt_t = max(M, ceil((NR_t − M)/I)×I + M).
PORrel_{t−L} = PORcpt_t
- PORrel_{t−L}: planned order release L buckets before receipt t [piece]; L: lead time in buckets [week or day; SI base is s]. Applies for deterministic, fixed lead time offsetting.
GR_child,t = Σ_parents ( PORrel_parent,t × usage_parent→child )
- Child gross requirement in t [piece]; usage is quantity of child per parent good unit [piece/piece]. Applies with 100% yield at parent; if parent has scrap s, plan parent receipts = GR_parent/(1−s) before this step.
Required receipts with scrap: PORcpt = GR_good / (1 − s)
- s: scrap fraction (0 ≤ s < 1). Units in piece; round up per lot rule. Applies where process yield losses occur before the point where “good” units are counted.
Q* = sqrt(2·D·S / H)
- EOQ. Q*: optimal lot size [piece]; D: demand rate [piece/s]; S: setup/order cost [₹/order]; H: holding cost rate [₹/(piece·s)]. Assumes constant D, instantaneous replenishment, no shortages, stationary costs.
POQ = Q* / d
- Period Order Quantity. POQ: number of periods covered [period]; d: average demand per period [piece/period]. Applies when converting EOQ to a time‑bucketed coverage rule.
Silver–Meal average cost over T periods: AC(T) = ( S + Σ_{t=1..T} H · (T−t) · d_t ) / T
- AC(T): average cost per period [₹/period] if one order now covers demand d_t in periods 1..T; choose smallest T with non‑increasing AC(T). Assumes holding cost linear in time and demand known.
Note: In practice we use weeks as buckets (common in MRP). SI base time is the second; 1 week = 604800 s if conversion is needed.
Worked examples
Example 1 — Direct netting with safety stock (single item, LFL) Data:
- Item X: lead time L = 1 week; lot size = LFL; safety stock SS = 10 piece.
- On‑hand at start PAB_0 = 20 piece.
- Scheduled receipt SR in week 2: 25 piece.
- Gross requirements GR by week 1..6: [30, 0, 40, 20, 0, 30] piece.
Steps:
- Week 1 net requirement
- Formula: NR_1 = max(0, GR_1 + SS − (PAB_0 + SR_1))
- Numbers: NR_1 = max(0, 30 + 10 − (20 + 0)) = max(0, 20) = 20 piece.
- LFL ⇒ PORcpt_1 = 20 piece; PORrel_0 = 20 piece (lead time 1 week).
- PAB_1 = PAB_0 + SR_1 + PORcpt_1 − GR_1 = 20 + 0 + 20 − 30 = 10 piece.
- Week 2 net requirement
- NR_2 = max(0, 0 + 10 − (10 + 25)) = 0 piece.
- PAB_2 = 10 + 25 + 0 − 0 = 35 piece.
- Week 3 net requirement
- NR_3 = max(0, 40 + 10 − (35 + 0)) = 15 piece ⇒ PORcpt_3 = 15; PORrel_2 = 15.
- PAB_3 = 35 + 15 − 40 = 10 piece.
- Week 4 net requirement
- NR_4 = max(0, 20 + 10 − (10 + 0)) = 20 ⇒ PORcpt_4 = 20; PORrel_3 = 20.
- PAB_4 = 10 + 20 − 20 = 10 piece.
- Week 5
- NR_5 = max(0, 0 + 10 − (10 + 0)) = 0; PAB_5 = 10.
- Week 6
- NR_6 = max(0, 30 + 10 − (10 + 0)) = 30 ⇒ PORcpt_6 = 30; PORrel_5 = 30.
- PAB_6 = 10 + 30 − 30 = 10 piece.
Answer:
- Planned order receipts: week 1: 20; week 3: 15; week 4: 20; week 6: 30 [piece].
- Planned order releases: week 0: 20; week 2: 15; week 3: 20; week 5: 30 [piece].
- Final PAB_6 = 10 piece. Assumes week‑0 release feasible; otherwise expedite or adjust time fence.
Example 2 — GATE‑level multilevel MRP with scrap and FOQ Product structure and data:
- End item P (make): L_P = 2 week; lot size LFL; SS = 0; on‑hand 0; no scheduled receipts. MPS (gross for P): week 6 = 50 piece; week 8 = 30 piece.
- A (make): usage P→A = 1; L_A = 1 week; LFL; SS = 0; on‑hand 0; no SR.
- B (make): usage P→B = 2; L_B = 1 week; LFL; SS = 0; on‑hand 0; no SR; process scrap s_B = 0.10 (10%).
- R (buy): used by A and B; L_R = 1 week; FOQ Q_R = 100 piece; SS_R = 20 piece; on‑hand 40 piece; no SR.
- C (buy): used by B; L_C = 2 week; FOQ Q_C = 50 piece; SS_C = 20 piece; on‑hand 40 piece; no SR.
- B’s BOM: A uses 2×R per A; B uses 1×R and 1×C per B.
Task: Compute planned order receipts and releases for all items through week 6; show R and C in detail.
Steps:
- Plan P from MPS (LFL)
- GR_P: w6 = 50, w8 = 30 [piece].
- LFL ⇒ PORcpt_P,w6 = 50; PORcpt_P,w8 = 30.
- Lead time offset: PORrel_P,w4 = 50; PORrel_P,w6 = 30.
- Explode to A and B gross requirements using parents’ releases (same buckets)
- A usage 1 per P ⇒ GR_A: w4 = 50; w6 = 30.
- B usage 2 per P ⇒ GR_B (good units): w4 = 100; w6 = 60.
- Plan A (L_A = 1, LFL, SS = 0)
- NR_A,w4 = 50 ⇒ PORcpt_A,w4 = 50; PORrel_A,w3 = 50.
- NR_A,w6 = 30 ⇒ PORcpt_A,w6 = 30; PORrel_A,w5 = 30.
- Plan B with 10% scrap (applies to receipts)
- Required receipts: PORcpt_B = GR_good / (1 − s_B).
- w4: PORcpt_B,w4 = 100 / 0.9 = 111.11… ⇒ 112 piece (round up, LFL integer).
- w6: PORcpt_B,w6 = 60 / 0.9 = 66.66… ⇒ 67 piece.
- Releases (L_B = 1): PORrel_B,w3 = 112; PORrel_B,w5 = 67.
- Explode to R and C using parents’ releases
- From A releases: PORrel_A,w3 = 50, w5 = 30.
- R for A: usage 2 ⇒ GR_R_from_A: w3 = 100; w5 = 60.
- From B releases: PORrel_B,w3 = 112, w5 = 67.
- R for B: usage 1 ⇒ GR_R_from_B: w3 = 112; w5 = 67.
- C for B: usage 1 ⇒ GR_C: w3 = 112; w5 = 67.
- Aggregate R gross: GR_R: w3 = 100 + 112 = 212; w5 = 60 + 67 = 127 [piece].
- From A releases: PORrel_A,w3 = 50, w5 = 30.
- Plan R (buy, L_R = 1, FOQ 100, SS_R = 20, on‑hand 40)
- Initialize PAB_R,0 = 40 piece.
- Week 1: GR = 0 ⇒ NR_1 = max(0, 0 + 20 − (40 + 0)) = 0 ⇒ PAB_1 = 40.
- Week 2: GR = 0 ⇒ NR_2 = 0 ⇒ PAB_2 = 40.
- Week 3: GR = 212 ⇒ NR_3 = max(0, 212 + 20 − 40) = 192 ⇒ FOQ 100 ⇒ PORcpt_R,w3 = ceil(192/100)×100 = 200; PORrel_R,w2 = 200. PAB_3 = 40 + 200 − 212 = 28.
- Week 4: GR = 0 ⇒ NR_4 = max(0, 0 + 20 − 28) = 0 ⇒ PAB_4 = 28.
- Week 5: GR = 127 ⇒ NR_5 = max(0, 127 + 20 − 28) = 119 ⇒ PORcpt_R,w5 = ceil(119/100)×100 = 200; PORrel_R,w4 = 200. PAB_5 = 28 + 200 − 127 = 101.
- Week 6: GR = 0 ⇒ NR_6 = max(0, 0 + 20 − 101) = 0 ⇒ PAB_6 = 101.
- Answer R: planned receipts: w3 = 200; w5 = 200. Planned releases: w2 = 200; w4 = 200. Final PAB_R,6 = 101 piece.
- Plan C (buy, L_C = 2, FOQ 50, SS_C = 20, on‑hand 40)
- Initialize PAB_C,0 = 40 piece. GR_C: w3 = 112; w5 = 67.
- Week 1: GR = 0 ⇒ NR_1 = max(0, 0 + 20 − 40) = 0 ⇒ PAB_1 = 40.
- Week 2: GR = 0 ⇒ NR_2 = 0 ⇒ PAB_2 = 40.
- Week 3: GR = 112 ⇒ NR_3 = max(0, 112 + 20 − 40) = 92 ⇒ PORcpt_C,w3 = ceil(92/50)×50 = 100; PORrel_C,w1 = 100. PAB_3 = 40 + 100 − 112 = 28.
- Week 4: GR = 0 ⇒ NR_4 = 0 ⇒ PAB_4 = 28.
- Week 5: GR = 67 ⇒ NR_5 = max(0, 67 + 20 − 28) = 59 ⇒ PORcpt_C,w5 = ceil(59/50)×50 = 100; PORrel_C,w3 = 100. PAB_5 = 28 + 100 − 67 = 61.
- Week 6: GR = 0 ⇒ NR_6 = 0 ⇒ PAB_6 = 61.
- Answer C: planned receipts: w3 = 100; w5 = 100. Planned releases: w1 = 100; w3 = 100. Final PAB_C,6 = 61 piece.
Summary answers:
- P: PORcpt w6=50, w8=30; PORrel w4=50, w6=30.
- A: PORcpt w4=50, w6=30; PORrel w3=50, w5=30.
- B: PORcpt w4=112, w6=67; PORrel w3=112, w5=67.
- R: PORcpt w3=200, w5=200; PORrel w2=200, w4=200; PAB_6=101.
- C: PORcpt w3=100, w5=100; PORrel w1=100, w3=100; PAB_6=61. Assumptions: Integer rounding up where needed; availability at bucket boundaries; no time fences blocking early releases; scrap applies to B’s receipts only.
Common mistakes
- Forgetting safety stock in netting or applying it to the wrong period (beginning vs end of bucket). Here we target end‑of‑bucket PAB ≥ SS.
- Not offsetting by lead time correctly; placing releases and receipts in the same bucket.
- Ignoring scheduled receipts when computing net requirements.
- Mishandling scrap: multiplying gross requirements by (1−s) instead of dividing planned receipts by (1−s).
- Lot sizing errors: rounding down FOQ or EOQ; forgetting minimum order quantities or increments.
- BOM level errors: not low‑level coding before the run, causing double counting.
- Treating EOQ as universally optimal in time‑phased, lumpy demand without checking heuristics (LFL, POQ, Silver–Meal, PPB).
For GATE ME
- Conceptual MCQs: definitions and differences among MPS, MRP, CRP, BOM, gross vs net requirements, receipts vs releases, low‑level code, pegging, regenerative vs net change, safety stock vs safety lead time.
- Short numericals: single‑item time‑phased record with given GR, on‑hand, scheduled receipts, SS, lead time, and a lot rule (often LFL or FOQ). Expect to compute PAB, NR, PORcpt, PORrel for 4–6 buckets.
- Multilevel explosion: small BOM (2–3 levels) with usage, lead times, SS, and either LFL or FOQ; possibly with scrap factors. Compute child gross requirements and plan purchases.
- Lot‑sizing cost comparison: EOQ/POQ or heuristic (Silver–Meal, PPB) over a few periods; identify the chosen horizon or lot multiple.
- Practice tips:
- Drill the standard MRP table columns and the netting formula with SS.
- Practise lead‑time offsetting and explosion carefully; always work parent releases → child gross.
- Be fluent with EOQ/POQ and at least one dynamic heuristic (Silver–Meal or PPB) for 3–4 periods.
- Check units and rounding; circle final releases (these drive downstream levels and suppliers).
Quick check
- In MRP, which drives a component’s gross requirements: the parent’s planned receipts or planned releases? Why?
- An item has PAB_0 = 15 piece, SS = 5 piece, SR in week 1 = 10 piece, GR in week 1 = 22 piece, LFL. Compute NR_1 and PORcpt_1.
- A parent needs 2 units of child per good parent; parent’s process has 20% scrap. For 100 good parents, how many child units are required at start?
Answers: 1) Planned releases, because components must be available when the parent order starts. 2) NR_1 = max(0, 22 + 5 − (15 + 10)) = 2 piece; PORcpt_1 = 2 piece. 3) Good parents required = 100; parents to start = 100/(1−0.2) = 125; child usage = 2 × 125 = 250 piece.
Finished this topic? Mark it so your progress, study plan and readiness keep up.
Stuck on something here?