Optimizing Inventory with EOQ Models

Introduction

In our last lesson, we explored how to make optimal capacity decisions in queueing systems by balancing the cost of service against the cost of waiting. We found the "sweet spot" that minimized total costs, whether by adjusting a single server's speed or by changing the number of servers. This principle of optimizing a trade-off between two opposing costs is a cornerstone of operations management.

Today, we will apply this same analytical logic to a different domain: inventory management. This lesson addresses the learning outcome: Calculate the optimal order quantity using the Economic Order Quantity (EOQ) model and its variants.

We will start by developing the classic EOQ model, which finds the ideal order size by balancing the costs of ordering inventory against the costs of holding it. We will then explore two crucial variations of this model that address more complex, real-world scenarios: the Production Order Quantity (POQ) model for items produced and consumed simultaneously, and the Quantity Discount model for when suppliers offer price breaks on larger orders.

Your background in applied mathematics and experience with optimization will be particularly relevant here, as the EOQ framework is fundamentally a cost minimization problem.

1. The Basic Economic Order Quantity (EOQ) Model

At the core of inventory management lies a fundamental trade-off:

• Ordering Costs (or Setup Costs): These are the fixed costs associated with placing an order, such as administrative expenses or shipping fees. If you place many small orders, your total annual ordering cost will be high.
• Holding Costs (or Carrying Costs): These are the costs of keeping inventory in stock, including storage, insurance, capital tied up, and risk of obsolescence. If you place few large orders, your average inventory level will be high, leading to high holding costs.

The goal of the EOQ model is to find the specific order quantity, $Q$, that minimizes the sum of these two costs.

To build our intuition for this model, let's start with a video that explains the underlying concepts, assumptions, and the graphical relationship between the costs.

Mathematical Formulation

As the video explained, we can express the total annual inventory cost as a function of the order quantity, $Q$. Let's define our variables:

• $D$ = Annual demand for the item (units/year)
• $S$ = Ordering (or setup) cost per order ($/order)
• $H$ = Holding (or carrying) cost per unit per year ($/unit/year)
• $Q$ = Order quantity (units/order)

The total annual cost, $TC$, is the sum of the annual ordering cost and the annual holding cost:

$TC(Q)=(\text{Annual Ordering Cost})+(\text{Annual Holding Cost})$

1. Annual Ordering Cost: This is the number of orders placed per year multiplied by the cost per order. The number of orders is $\frac{D}{Q}$.
   $\text{Annual Ordering Cost}=\frac{D}{Q}S$

2. Annual Holding Cost: This is the average inventory level multiplied by the holding cost per unit. The average inventory is $\frac{Q}{2}$.
   $\text{Annual Holding Cost}=\frac{Q}{2}H$

Combining these gives the total cost function:
$TC(Q)=\frac{DS}{Q}+\frac{HQ}{2}$

To find the order quantity $Q^{\ast }$ that minimizes this cost, we can take the first derivative of $TC(Q)$ with respect to $Q$, set it to zero, and solve for $Q$.

$\frac{d(TC)}{dQ}=-\frac{DS}{Q^2}+\frac{H}{2}=0⟹\frac{H}{2}=\frac{DS}{Q^2}$

Solving for $Q$ gives us the famous EOQ formula:

$Q^{\ast }=\sqrt{\frac{2DS}{H}}$

The next video segment will walk you through this derivation and a simple calculation.

Worked Examples

To solidify your understanding, let's work through two practical examples that cover common scenarios.

As you saw in the first example, ordering a quantity other than the EOQ results in a higher total cost, confirming that $Q^{\ast }$ is indeed the minimum. The second example demonstrated a crucial practical point: holding cost ($H$) is often not a given number but is driven by the value of the inventory itself. Calculating it as H = I * C, where I is the annual holding cost rate and C is the unit cost, is standard practice.

2. Variants of the EOQ Model

The basic EOQ model is powerful, but its assumptions (like instantaneous delivery and no price discounts) are restrictive. We will now explore two important variants that relax these assumptions to better model real-world situations.

Variant 1: Production Order Quantity (POQ) Model

The POQ model (also known as the Economic Production Quantity or EPQ model) is used when inventory is replenished gradually over time, as in a manufacturing setting where a company produces its own parts.

In this scenario, items are being produced and added to inventory while simultaneously being used or sold. This changes the inventory level dynamics.

The key variables are:

• $p$ = daily production rate
• $d$ = daily demand (or usage) rate

During the production run, inventory accumulates not at the rate $p$, but at the rate $p-d$. This leads to a lower maximum inventory level compared to the basic EOQ model, which in turn means lower average holding costs.

This modification results in a slightly different formula for the optimal production quantity, $Q_p^{\ast }$:

$Q_p^{\ast }=\sqrt{\frac{2DS}{H(1-d/p)}}$

Notice the factor $(1-d/p)$ in the denominator. Since $d<p$ for inventory to accumulate, this factor is less than 1, making the denominator smaller and the resulting optimal quantity $Q_p^{\ast }$ larger than the basic EOQ. This makes intuitive sense: because average holding costs are lower in this model, it is more economical to produce in larger batches.

Variant 2: Quantity Discount Model

This is a very common scenario where suppliers offer a lower per-unit cost for larger order quantities. This complicates the analysis because the annual purchase cost, which we previously ignored as a constant, now varies with the order quantity. The holding cost $H$ might also change if it's based on the unit cost.

The total cost function must now include the cost of the product itself:
$TC(Q)=\frac{DS}{Q}+\frac{HQ}{2}+DC$
where $C$ is the per-unit cost.

Since $C$ (and possibly $H$) changes at different quantity thresholds (price breaks), we cannot find a single optimal $Q^{\ast }$ with one formula. Instead, we use a systematic process:

1. Calculate $Q^{\ast }$ for each price level: Starting with the lowest unit cost, calculate the EOQ using the corresponding $H$.
2. Check for feasibility: For each calculated $Q^{\ast }$, determine if it falls within the quantity range required to obtain that price.
3. Adjust if necessary:
   • If $Q^{\ast }$ is feasible, it is a candidate for the optimal order quantity.
   • If $Q^{\ast }$ is too low to qualify for the price, adjust the candidate order quantity up to the minimum required for that price break.
   • If $Q^{\ast }$ is too high, that price level is not relevant for the EOQ calculation at that tier.
4. Calculate Total Cost: Compute the total annual cost (including purchase cost) for each candidate quantity identified in the previous steps.
5. Select the Optimum: The order quantity with the lowest total annual cost is the overall optimal choice.

Conclusion

Key Takeaways:

• The Economic Order Quantity (EOQ) model provides a quantitative basis for inventory decisions by minimizing the sum of ordering and holding costs. The optimal quantity is found at the point where these two costs are equal.
• The basic EOQ formula is $Q^{\ast }=\sqrt{\frac{2DS}{H}}$. It operates on a set of simplifying assumptions, such as constant demand and instantaneous replenishment.
• The Production Order Quantity (POQ) model is a variant for scenarios with simultaneous production and consumption. It results in a larger optimal batch size due to lower average inventory levels.
• The Quantity Discount model provides a systematic algorithm to determine the optimal order quantity when unit price varies with order size. This requires calculating and comparing the total cost (including purchase cost) at each price break.

Next Steps

In this lesson, we've focused on a highly quantitative and specific optimization problem. In our next lesson, we will broaden our perspective to a more holistic framework for operational excellence. We will explore the five core principles of lean management and learn to identify the eight forms of waste (Muda) in a process. This will shift our focus from mathematical models to a powerful philosophy for continuous improvement.