JURNAL OPTIMASI SISTEM INDUSTRI - VOL. 18 NO.

2 (2019) 116-124

Available online at : http://josi.ft.unand.ac.id/

Jurnal Optimasi Sistem Industri

| ISSN (Print) 2088-4842 | ISSN (Online) 2442-8795 |

Case Study

Analysis of Inventory Control by Using Economic Order Quantity Model – A Case

Study in PT Semen Padang
Prima Fithri, Alizar Hasan, Fadhita Maisa Asri
Department of Industrial Engineering, Faculty of Engineering, Universitas Andalas, Kampus Limau Manis, Padang 25163, Sumatera Barat, Indonesia

ARTICLE INFORMATION A B S T R A C T

Received: February 2, 19 Inventory control is a very important issue. It is because the amount of inventory will determine
Revised: October 30, 19 or affect the smoothness of the production process as well as the effectiveness and efficiency of
Available online: October 31, 19 the company. PT Semen Padang is a manufacturing company that produces 10,400,000 tons of
cement per year. The achievement of the cement production target at this company depends on
the availability of raw materials needed in the cement production process itself. Gypsum is an
KEYWORDS additional material of the cement production process which is very important because it is a raw
material that must exist in the process of making cement. So, if the inventory of gypsum cannot
PT Semen Padang, inventory control, EOQ, meet the needs of production, then the production process of cement making will be disrupted.
cement, Bulk Material Gypsum PT Semen Padang is using the Min-Max method for the inventory control. But the costs are
quite high. The cost of inventory can be minimized by using another method such as EOQ
CORRESPONDENCE (Economic Order Quantity). The conclusions of this research are in 2016 by using EOQ method,
the optimal order quantity is 32,073 ton per order, and the frequency is 9 times in a year with
total cost Rp. 4,757,673,813.48, and in 2017, the optimal order quantity is 34,856 tons per order
Phone: +628116644988
and the frequency is 9 times in a year with total cost Rp. 9,694,805,608.36.
E-mail: [email protected]

Gypsum is one of the raw materials that is bought from the

INTRODUCTION suppliers that located in Cilacap, Central Java. The purpose of
adding gypsum is to slow the cement hardening process. Gypsum
Inventory control is a very important issue [1–3]. The amount of is an additional material that is very important because without
inventory will determine or affect the smoothness of the this material the cement production process cannot be done. So,
production process as well as the effectiveness and efficiency of if the inventory of gypsum does not meet the needs of production,
the company [4]. Delays or out of inventory will not only then the production process of cement will be disrupted. The
influence the production process but also cause a loss of demand for gypsum in PT Semen Padang is presented in Figure
opportunity to increase profits from a company. The amount or 1.
level of inventory required by a company is different for each 25000
company, depending on the quantity of production, the capacity
of the plant and the production process [5,6]. 20000
Demand

15000
PT Semen Padang is a manufacturing company that produces
cement. The company can produce 10,400,000 tons of cement per 10000
year. The achievement of the cement production target at this
company depends on the availability of raw materials needed in 5000
the cement production process itself. The raw materials for
0
making cement are limestone, silica, clay, iron sand or copper
1 3 5 7 9 11 13 15 17 19 21 23
slag, gypsum and pozzolans. Limestone and silica are raw Period
materials that produced in the company, while other raw materials
are bought from the suppliers. Figure 1. Demand of Gypsum in PT Semen Padang

As is shown on Figure 1, the used of gypsum in PT Semen Padang

is fluctuate and keep increasing in period 22nd. The availability

DOI: 10.25077/josi.v18.n2.p116-124.2019 Attribution-NonCommercial 4.0 International. Some rights reserved

 FITHRI ET AL. / JURNAL OPTIMASI SISTEM INDUSTRI - VOL. 18 NO. 2 (2019) 116-124

of gypsum determines the cement production process at PT 1. Qualitative Forecasting Methods

Semen Padang. Therefore, it is necessary to control the inventory Forecasting that combines an intuitive, emotional, personal
of gypsum so the production process will be run smoothly. experience, and value systems decision-makers to predict. This
method is usually used when there is no or little of available
PT Semen Padang is already using a method for their inventory past data. Qualitative methods are widely known it is the
control that is the Min-Max method. This method is used to Delphi method and the nominal method (nominal group
determine the minimum and maximum stock. The used of technique).
determining the minimum stock is to avoid a shortage of raw 2. Quantitative Forecasting Methods
material, and for the maximum stock is to avoid waste of storage Using a variated mathematical model with past data and causal
cost. But by using this method, the costs incurred by the company variables to forecast demand. Quantitative methods consist of:
can be quite high. The cost of inventory can be minimized by Methods of Time series ("Hands-free"/graphics, moving
using another method such as EOQ (Economic Order Quantity) average, weight moving average, exponential smoothing,
[7–10]. In this report, there will be a calculation to get the total simple linear regression, interpolation Gregory-Newton,
inventory cost by using EOQ and compare it with the inventory winter, ARIMA), and Methods "Non-time Series" (Structural
cost which has been applied by the company. So, the company Models).
will be able to minimize the inventory cost.
Quantitative forecasting can be applied when there are three of
Based on the background described above, the problem the following conditions:
formulation that to be discussed is how to control the inventory 1. Available of information about the previous period.
of raw material gypsum for the next period and whether the cost 2. Such information can be quantified in the form of numerical
of inventory that has been set by the company is already optimal. data.
The objective of this research is to determine the optimal 3. It can be assumed that some aspects of the pattern of the past
inventory cost for raw material gypsum by using EOQ. will continue in the future.

The scope of the research that conducted in PT Semen Padang are a. Exponential methods
as follows: This method did weighting method decreases exponentially
1. Inventory control is done for raw material gypsum toward the value of a variable or observation ago. Each
2. Forecasting is based on the historical data of gypsum demand observation data has contributed to determining the value of
in 2014 and 2015 for the next two years. forecasting the previous period. However, the calculation is
fairly represented by observational data and the results of
LITERATURE REVIEW forecasting in the last period.
b. Quadratic method
Forecasting The quadratic method is a non-linear trend, and if it drawn
shaped like curve lines. This method is usually used or applied
Forecasting is an activity or business that is used to predict the to historical data which, if drawn will form lines are not straight
future state through state testing in the past while forecasting or parabolic.
demand is the level of demand for the products that are expected c. Trend cyclical method
to be realized for a certain period in the future [11,12].
Forecasting process will get the results used by management Definition of Inventory
forecasting production or operation in making decisions Inventory is materials, supplied parts and materials in the process
regarding the selection process, capacity planning, and layout of contained in the company's production process, also finished
the facility, and also for decisions that are constantly about goods or products supplied to fulfill the demand of customers at
planning, scheduling, and inventory. Demand forecasting is any time. So, it can be concluded that inventory is the availability
useful to know the pattern of customer trends and information of any stock or resources used in an organization that is stored for
about the product use in the future period to meet certain objectives.

The purpose of forecasting in operational management is to An inventory control that is run by a company certainly has
reduce uncertainty in production, so that proactive certain goals, as follows:
measures/anticipatory can be done, and for purposes of 1. To fulfill the needs or the demands of the customer faster (to
production scheduling. Forecasting can be affected by the satisfy the customer).
external environment and internal environment. The external 2. To maintain the continuity of production or keep the company
environment can be either income customer, promotion of did not get the stock-outs that can affect to the problem of the
competitors, competitors' prices, availability of products, production process such as the stopping of the production
competitive effectiveness and efficiency of channels used, the process, this is because of the reasons:
characteristics of customers, and so on while the internal a. The possibility of goods (raw and supporting materials)
environment is the policies carried out within the company, such become scarce, so it is difficult to obtain.
as promotion policies, costs, and channels. b. The possible of suppliers that can be late to send the goods
ordered.
Forecasting Methods 3. To maintain and if possible to increase the company's sales and
profits.
There are two kinds of forecasting methods, as follows:
DOI: 10.25077/josi.v18.n2.p116-124.2019 Fithri et al. 117
 FITHRI ET AL. / JURNAL OPTIMASI SISTEM INDUSTRI - VOL. 18 NO. 2 (2019) 116-124

4. To keeping small purchases can be avoided that can increase 2. Inventories of parts of products purchased (Purchased
the ordering cost higher. parts/components stock), is the inventory of goods which
consist of parts received from other companies, which can be
assembling directly with other parts without going through the
Function of Inventory production process before.
Inventory exists because supply and demand are difficult to
synchronize perfectly. Inventory enables a company to support 3. Inventories of auxiliaries’ materials or the supporting goods are
the customer’s services, logistics or manufacturing activities in a the inventory of materials which required in the production
situation where purchasing or manufacturing of the items is not process to help the success of the production or used in the
able to satisfy the demand [13]–[16]. Not only in manufacturing operation of a company, but not a part or component of the
sectors, inventory problem can be found in services area like finished goods.
hospital [17]. Customer desire has always been a vital issue in a
company not only to maintain sales but also to increase it. One of 4. Inventories of finished goods in the process (work in
the functions of inventory is the anticipate the fluctuate demand process/progress stock), is the inventory of goods coming out
so the company can optimize the ability to fulfill the customer of every part in a single plant or materials which have been
demand to increase the company’s performance. Inventory has processed into a form but needs more to be reprocessed for
several functions to create stability in the company's operations process becoming the into finished goods. But maybe semi-
activity and hide the operational problem. finished goods for a factory is a finished good for other
factories because the production process ends there. May also
Types of Inventory semi-finished goods that are the raw material for other
The types of inventory in the company can be categorized in companies who will process it into finished goods.
several ways. Based on the function, inventory can be divided
into: 5. Finished goods inventory is the inventory of goods that have
been fully processed or processed in the factory and ready for
1. Batch Stock or Lot Size Inventory, is the inventory that appears sale to customers or other companies. So this is a finished
where the goods purchased, goods made or transported in large product and has been prepared for sale.
quantities, so the goods gained more and faster than the used
or expenditure, and for a while can create an inventory. The The Function of Inventory
profits can earn with the Batch Stock or Lot Size Inventory
include: Inventory has several functions to create stability in the
a. Getting a discount on the purchase price. company's operations activity. The functions of inventory are as
b. The efficiency of production follows:
c. There is saving cost for the handling cost.
1. Decoupling function.
2. Fluctuation Stock is inventory that use to face the fluctuations The important function of inventory allows the operations of
of customer demand that cannot be predicted. In this case, the internal and external companies have freedom. This decouples
company will make preparations to be able to fulfill customer inventories allows the company to fulfill the customer demand
demand, if the level of demand shows irregular or not fixed and without depending on the supplier.
fluctuations in demand cannot be predicted before. So if there
are a very large demand fluctuation, then this inventory 2. Economic Lot Sizing function
(fluctuation stock) takes a very big too to keep the possibility Economic Lot Sizing function is a function that enables the
of the rise and fall of demand. company to produce and purchase the number of resources that
can reduce the costs per unit. Lot Size function is necessary to
3. Anticipation Stock is the inventory which supplies are held to consider cost savings. The savings from the discount of
face the unpredictable fluctuations, based on seasonal patterns purchasing, transportation costs, and so on. These savings costs
contained within one year and to deal with the use or sale of occur because the company bought in larger quantities.
the increased demand. Anticipation Stock is also intended to
keep the possibility of difficulty obtainable materials, so it’s 3. Anticipation function
not to interrupt the production process or can avoid the delay Anticipation function as the inventory to anticipate the demand
of production. that can be forecast and keep the possibility of difficulties in
obtaining the raw materials. This function to cope with the
The inventory also can be grouped based on types and position of uncertainty of the delivery time and the receipt of raw materials
the item in the sequence of the work product that follows: during the period of re-ordering. This function is very
1. Inventories of raw materials are the inventory of tangible goods important to maintain of smoothing production process
used in the production process, where goods can be obtained
The Inventory Costs
from the natural source or purchased from suppliers or
companies that produce raw materials for the company. The Inventory costs are all the expenses and losses caused by the
raw materials in factories use to process become finish goods inventory activities. Cost of inventory is generally divided into
through of some process. four types:
1. Purchasing Cost

118 Fithri et al. DOI: 10.25077/josi.v18.n2.p116-124.2019

 FITHRI ET AL. / JURNAL OPTIMASI SISTEM INDUSTRI - VOL. 18 NO. 2 (2019) 116-124

The purchasing cost is the expenses cost that used to the 3. Keeping small purchases can be avoided because this would
purchase of goods with a certain number depending on the make the ordering cost higher.
number of items purchased and the price per unit of goods. The
purchasing cost becomes very important when the price of The importance of inventory control about the production process
goods is influenced by the size of the purchasing with the price is to ensure a smoothing production process, and the continuity
discount is existing where the price per unit will decrease if the of the plant do not let the company stock out of inventory so that
number of purchases increased, and also the opposite. the production process can continue smoothly.

2. Procurement Cost
Economic Order Quantity (EOQ)
The procurement costs are the costs related to the purchasing The model of economic order quantity (EOQ) is a mathematical
of goods that consist of ordering cost if the goods necessary model to determine the number of items needs to be ordered to
come from outside the company. Ordering cost is all expenses fulfill projected demand with the cost of inventory is minimized.
caused by the activities bring in goods from outside, such as EOQ aims to determine the reorder point for independent
the cost determining of the supplier, typing orders cost, demand, which the demand for a component that does not depend
shipping cost, the transportation cost, inspection cost, packing on the other components of demand. The model used in
cost, telephone charges, and other costs. Ordering cost is calculating the EOQ is:
assumed constant for each time of reservation.
2𝐴𝐷
𝐷 𝐸𝑂𝑄 = 𝑄 ∗ = √ (3)
𝑂𝑟𝑑𝑒𝑟𝑖𝑛𝑔 𝐶𝑜𝑠𝑡 = 𝑥(𝐴) (1) ℎ
𝑄

3. Carrying Cost/Holding Cost

The holding cost is all the expenses caused by the activities of
saving goods in a certain period.

𝑄
𝐻𝑜𝑙𝑑𝑖𝑛𝑔 𝐶𝑜𝑠𝑡 = 𝑥(ℎ𝑥𝐶) (2)
2

4. Stock out Cost/Shortage Cost

The stockout cost reflected the economic consequences caused
by stock out of inventory. This could be disadvantages for the
company because it would make the company's production
process become slow down and the company of the opportunity Figure 2. Classical Inventory Model
to make a profit or a customer can move to another company
because the demand is not fulfilled. This can make affected to Figure 3 shows economic order quantity graph. It consists of
the company's image in the customer perception. As for that holding costs, total cost, and ordering costs.
including the cost of a stockout is the number of items that are
not fulfilled, the fulfillment time and procurement costs of
emergencies.

The Inventory Control

Inventory control is an activity to determine the level and
composition of assembly (parts) inventory component, raw
materials and finish good/products so that the company can
protect the smoothing production process and sales also the
learning needs of companies with effective and efficient [18].
Inventory control activities are not limited to the determination
on the level and composition of inventory, but also including the Figure 3. Economic Order Quantity Graph
regulation and supervision or the procurement of the necessary
ingredients according to the amount and time required. The total cost of inventory with reservations on EOQ:

If the raw material inventories are low or empty it will lead to Total Cost = Ordering Cost + Holding Cost
delays in the production process and even result in the stopping = The cost each time order x order frequency +
of the production process, but conversely, if too much inventory holding cost/unit/period x number of average
will also lead to waste. The purpose of control is expressed as an inventory
effort to [18]: = (𝐴𝑥𝐹) + (ℎ𝑥𝑚)
1. Keeping the company to not to run out of inventory that can 𝐴𝐷 ℎ𝑄
= 𝑄
+ 2
(4)
make stopping of the production process.
2. Keeping the formation of inventories by companies is not too
great or excessive, so the costs for inventory of raw materials is Assumptions to apply the EOQ model are:
not too high. 1. The demand for the product is constant, uniform and
outermost (Deterministic).
DOI: 10.25077/josi.v18.n2.p116-124.2019 Fithri et al. 119
 FITHRI ET AL. / JURNAL OPTIMASI SISTEM INDUSTRI - VOL. 18 NO. 2 (2019) 116-124

2. Price per unit of product is constant.

3. The storage cost per unit per year (H) is constant. Data required to solve the problems of requirement planning and
4. The ordering cost per order (S) is constant. inventory control in PT Semen Padang are consist of primary data
5. The time between orders is done and goods received (lead and secondary data.
time, L) is constant. a. Primary Data
6. There is no shortage of goods or return orders Primary data is the data that taken from the interview. The first
interview is done with the head of Procurement Planning and
Control Division about the cost related to procurement of
Safety Stock gypsum in PT Semen Padang such as ordering cost, holding
Safety Stock is the amount of finished goods inventory which is cost and the lead time ordering gypsum. The second interview
also referred to as a buffer stock. Safety stock is used to fulfill the is done with the head of Inventory Planning and Control
demand of customer when uncertainty things happen suddenly. The Division and Employee of Production Planning and Control
equation for calculating safety stock are: Division about the importance of gypsum in PT Semen Padang.
The last interview is done with the employee of the Production
Safety stock = safety factor x standard deviation (5) Planning and Control Division about the purchasing cost of
gypsum.
Standard deviation is the result of calculations that use the demand
data during the period. The formula for calculating standard b. Secondary Data
deviation: Secondary Data is the data from the Production Planning and
Control Division in PT Semen Padang, the list of the data is as
follow.
∑(𝑥−𝑥̅ )2
𝑆=√ 𝑛
(6) 1. Historical Data Demand of Gypsum in 2014 and 2015
2. Purchasing Cost of Gypsum
3. Ordering Cost of Gypsum
The safety factor is a calculation of a safety factor that depends
4. Holding Cost of Gypsum
on the service level provided. In determining the service level,
some companies specify different service levels for some types or
Historical data demand of gypsum is used to forecast the demand
categories of goods, for example:
of gypsum for the next 24 months. Then determine the number
of optimal ordering based on methods and calculate the inventory
Raw material: 100% service level
cost. The result is compared with the inventory cost that the
Goods program: 100% service level
company spends in PT Semen Padang to know whether the
inventory cost is already optimal or not. The analysis is conducted
Raw materials for:
on the data processing that has been done and refers to literature
Vital equipment: The service level of 99%
found on a theoretical basis.
Essential equipment: The service level of 95%
Auxiliary equipment: The service level of 90%
Operational equipment: 85% service level
RESULTS AND DISCUSSION

Data collection was done during the implementation of industrial

METHOD
training in PT Semen Padang. The data needed for the calculation
are as follows. Data of gypsum historical demand PT Semen
The research methodology is derived from steps that carried out
Padang in the year 2014 and 2015 can be seen in Figure 1.
in the research—the steps starting from conducting the
preliminary review, literature review, problem identification, Purchasing Cost of Gypsum
problem formulation, data collecting, data processing, and
The purchasing cost of gypsum is 25 USD per ton. In rupiah the
analysis.
purchasing cost is Rp. 333,350.00 per ton. The cost of gypsum
per year can be seen in Table 1.
Problem observations were conducted to determine and observe
real conditions that occur in related companies. The observation
Table 1. Purchasing Cost of Gypsum
was also followed by interviews with the head of Procurement
Planning and Control Division, the head of Production Planning Year Purchasing cost
and Control Division and employees of Production Planning and 2014 Rp. 68,005,760,784.70
Control Division. 2015 Rp. 75,931,115,699.30

The problem is the stock of gypsum in PT Semen Padang. Ordering Cost of Gypsum
Sometimes PT Semen Padang is run out of gypsum stock, and
that is dangerous because gypsum is the most important additive Estimation of ordering cost is 3 % of the purchasing cost per year.
material for making cement. Without gypsum, the production It is because the purchasing cost of gypsum is the cost until the
process of cement will not be run smoothly. The formulation of gypsum arrives at the storage. So, the ordering cost that consists
the problem based on the identification of the problem is how to of transportation cost and the administration cost is included in
control the inventory of gypsum for the next period and is the cost the purchasing cost. The estimation of ordering cost can be seen
of inventory that the company’s spend is already optimal. in Table 2.

120 Fithri et al. DOI: 10.25077/josi.v18.n2.p116-124.2019

 FITHRI ET AL. / JURNAL OPTIMASI SISTEM INDUSTRI - VOL. 18 NO. 2 (2019) 116-124

Table 2. Ordering Cost of Gypsum

Year Ordering cost Table 4. Recapitulation of Error Calculation
2014 Rp. 170,014,401.96 No. Forecasting Method MAPE Minimum
2015 Rp. 189,827,789.25
1 Regression Linear 0.1092
2 Cyclic 0.1598
Holding Cost of Gypsum 0.1053
3 Exponential 0.1053
Holding cost is the estimation cost from the actual cost because
4 Trend Cyclic 0.106
the company did not know the exact number of the gypsum
holding cost every year. The estimation of holding cost can be
seen in Table 3. Forecasting Verification for the Selected Method

Table 3. Holding cost of gypsum The purpose of verification is because there is a possibility of data
forecasting results that outside of the control limits. Data which
Year Holding Cost
are outside the control limits should be discarded and do re-
2014 Rp. 98,719.60
forecasting again.
2015 Rp. 98,223.48

Based on the first verification result there are 2 data that out of
Lead Time Ordering Gypsum the control limit that should be discharged. Then do forecasting
with the 22 data with the chosen forecasting method and
Lead Time is the interval of time between the ordering time with
continued to the second verification.
the arrival time of the material. Lead Time ordering gypsum is 30
days. The process to know the optimal inventory cost by using
Based on the second verification result shows that there is no data
EOQ method is done by following steps below.
outside the control limits, so the results can be used to forecast
Plot Data of Gypsum Historical Demand the demand for the next period. The result of forecasting using
exponential method can be seen in Figure 4. Based on the forecast
The first step for forecasting is to plot the data of historical demand, then the inventory control was calculated using EOQ
demand. The purpose of the plot data is to see the pattern of the method.
data to choose the right method that suit the pattern. The plot data
of historical demand gypsum in PT Semen Padang shows in The year 2016:
Figure 1.
2AD
EOQ =√
h
Forecasting Method
2 x Rp 179,921,095.61 x 281,505.64 ton
There are for methods that used for the forecast in this solving =√
Rp 98,471.54
problem, that are linear regression, exponential, cyclical and = 32,073 ton
cyclical trends.
D
The Calculation of Forecasting Method Error Frequency= Q
281,505 ton
Forecasting method error is calculated using the Mean Absolute =
32,073 ton
Percentage Error (MAPE) method. The error of forecasting = 8.78
method with the smallest error is the method chosen to perform = 9 times
forecasting next 12 periods. Recapitulation of error calculation of
each method can be seen in Table 4. Year 2016:

Based on the recapitulation calculation error result, exponential Inventory Cost

forecasting method chosen as the forecasting method for next = Ordering Cost + Holding Cost
period forecast, because this method has the minimum error based = (9 x Rp 179,921,095.61) + ((16,036.67 ton + 15,834.303 ton) x
on MAPE calculation (the smallest error). Rp 98,471.54)
= Rp 4,757,673,813.48
Table 5. The result of EOQ method
2014 2015 2016 2017
Ordering cost Rp 170,014,401.96 Rp 189,827,789.25 Rp 179,921,095.61 Rp 179,921,095.61
Demand 204,007.08 227,781.96 281,505.64 332,47.23
Holding cost Rp 98,719.60 Rp 98,223.48 Rp 98,471.54 Rp 98,471.54
Lot size order 26,508 29,672 32,073 34,856
Frequency 8 8 9 10

DOI: 10.25077/josi.v18.n2.p116-124.2019 Fithri et al. 121

 FITHRI ET AL. / JURNAL OPTIMASI SISTEM INDUSTRI - VOL. 18 NO. 2 (2019) 116-124

Table 6. The Inventory Cost of EOQ Method

Year Ordering cost Holding cost Inventory cost
2014 Rp 1,360,115,215.69 Rp 11,305,769,680.83 Rp 12,665,884,896.5
2015 Rp 1,518,622,313.99 Rp 8,196,735,197.01 Rp 9,715,357,510.9 Year 2016:
2016 Rp 1,619,289,860.45 Rp 3,138,383,963.04 Rp 4,757,673,813.5 Max Inventory
2017 Rp 1,799,210,956.05 Rp 7,895,594,652.31 Rp 9,694,805,608.4 = Safety Stock + EOQ
= 15,834.303 ton + 32,073 ton
Safety Stock = 47,908 ton
The year 2016:
2 Year 2017:
∑ii=1 (y' -y̅' )
Standard Deviation =√ Max Inventory
n-1

= 1,122.69 = Safety Stock + EOQ

= 62,753.454 ton + 34,856 ton
Table 7. The Standard Deviation for Each Year
= 97,610 ton
Year Standard deviation
2014 7,180.31 Inventory Cost in PT Semen Padang
2015 4,864.90
To compare the inventory cost in the company and by using EOQ
2016 1,122.69
method, we must know both of inventory cost. Because of the
2017 4,449.38
inventory cost of PT Semen Padang is classified, there must be an
estimation to know the number of total inventory cost. This
Lead Time = 30 days
company is using the Min-Max method to control the inventory
Service Level (Z) = 100% - z
of gypsum, so the estimation of the inventory cost is calculated
= 100% - 99%
using the Min-Max method.
= 2.5750 (From Normal Curve Table)
Year 2014:
Year 2016:
Safety Stock
Safety Stock
= (Maximum used – Average of used) x Lead time
= Z x Standard Deviation x √Lead time
= (22,164 ton – 17,001 ton) x 1 month
= 2.5750 x 1,122.69 x √30
= 5,163.51 ton
= 15,834.303 ton
Min stock
Table 8. The result of safety stock
= (Average of used x Lead time) + Safety stock
Year Safety stock (ton) = ( 17,001 ton x 1 month) + 5,163.51 ton
2014 101,270 = 22,164.10 ton
2015 68,613.85
Max stock
2016 15,834.30
= 2 x (Average of used x Lead time)
2017 62,753.45 = 2 x ((17,001 ton x 1 month)
= 34,001.18 ton
Reorder Point
Year 2016: Reorder Point
Reorder Point = Max stock – Min stock
= Safety Stock x (Lead time x Average used) = 34,001.18 ton – 22,164.10 ton
= 15,834.303 ton x (1 month x 23,458.80 ton) = 11,837.08 ton
= 39,293.10581 ton
Inventory Cost
Year 2017: = ((Total used/Reorder point) x Ordering cost) + (Holding cost x
Reorder Point Total used)
= Safety Stock x (Lead time x Average used) = ((204,007.08 ton / 11,837.08 ton) x Rp 170,014,401.96) + (Rp
= 62,753.454 ton x (1 month x 27,706.02 ton) 98,719.60 x 204,007.08 ton)
= 90,459.47317 ton = Rp 23,069,623,760.92
Maximum Inventory

Table 9. The Result of Using the Min-Max Method

2014 2015 2016 2017
Safety stock 5,163.51 2,952.73 1,830.18 2,161.53
Minimum stock 22,164.10 21,934.56 25,288.98 29,867.55
Maximum stock 34,001.18 37,963.66 46,917.61 55,412.04
Reorder point 11,837.08 16,029.10 21,628.62 25,544.49
Inventory cost Rp. 23,069,623,760.92 Rp. 25,071,089,876.70 Rp. 30,062,043,061.22 Rp. 35,080,802,308.82

122 Fithri et al. DOI: 10.25077/josi.v18.n2.p116-124.2019

 FITHRI ET AL. / JURNAL OPTIMASI SISTEM INDUSTRI - VOL. 18 NO. 2 (2019) 116-124

Table 10. Comparison of Inventory Cost

Inventory cost of the Inventory cost using
Year Difference Presentation
company EOQ
2014 Rp. 23,069,623,760.92 Rp. 12,665,884,896.52 Rp. 10,403,738,864.39 45.10%
2015 Rp. 25,071,089,876.70 Rp. 9,715,357,510.99 Rp. 15,355,732,365.70 61.25%

Figure 4 shows that the demand forecasting is getting higher for

Comparison of Inventory Cost of Company and Using each period. This is because the selected method that is used to
Methods forecast the demand is the Exponential Method. Exponential
The comparison between the inventory cost of the company and method graph is always getting higher for the next period because
the inventory cost using method can be seen in Table 10. The it used for the trend data.
analysis is done based on the results of data processing that has
been done before. The first step to forecast demand is to plot the After getting the forecast demand, the inventory control of
data historically. The purpose of the plot data is to see the pattern gypsum was calculated to meet the objective of the problem-
to determine the right method for forecasting the demand. Based solving. The method of inventory control that is used in this
on the plot data, forecasting methods that we can use are the problem is EOQ. From data calculation, the result of gypsum
Linear Regression method, cyclical method, Exponential method, usage based on production demand is 281,505.64 ton in 2016.
and Cyclical Trends method. Based on EOQ method, the value of EOQ is 32,073 ton. It means
that the optimal amount of ordering gypsum is 32,073 ton for
Forecast error is the difference between the actual demand and every order. The frequency is nine times. It means the optimal
forecast demand. This is useful when you want to determine if frequency that the company order gypsum is nine times in the
one forecasting method is better than another. From this problem, year 2016. And the total inventory cost is Rp. 4,757,673,813.48.
we used four different methods to determine the forecast demand The amount of safety stock to be saved by the company as a level
for gypsum in the next period. So, the calculation of forecast error of extra stock that is maintained to mitigate the risk of stock outs
is to determine the best forecast method to use for forecasting the or due to uncertainties in supply and demand is 15,834.303 ton.
demand of next period. The method that is used to calculate the The reorder point, where the company must start to order gypsum
forecasting error in this problem is MAPE. In the calculation, the is when the gypsum stock is in 39,293.10581 ton. The maximum
selected method is an Exponential method that has MAPE value inventory in the storage is 97,610 ton.
0.1053.
PT Semen Padang is using Min-Max method for the inventory
The next step after got the result of forecasting using the selected control. Because of the inventory cost is classified, we used the
method is verification of the data. The purpose of the verification min-max method calculation for the estimation of the company
is to see whether the forecast demand interprets the actual demand spend on inventory cost of gypsum. Based on the calculation in
based on the historical data of the company or not. The method 2014, the inventory cost of the company is Rp. 23,069,623,760.92
that is used to verify the forecast model is the Moving Range and the inventory cost by using EOQ is Rp. 12,665,884,896.52, it
Chart. has a difference about 45%. The purpose of EOQ is the order of
quantity that minimizes the total of holding cost and ordering cost
Based on the result of 1st verification, from the 24 data of forecast for the year. So, if the company using EOQ as a method for
demand, 2 data are deleted because the residual value is outside inventory control the company will save 45% of the total cost that
the control limit, there are the 4th and 7th. For the 2nd the company spends in that year. It is also the same in the year
verification, all the data is inside the control limit. So, no more 2015, the difference between the total costs in the company and
data has to be thrown away, and the verification step is done. using methods is 61%.
Then, forecast the next demand period by using the 22 data of the
forecast data. The result of forecasting can be seen in Figure 4. ACKNOWLEDGMENT

Authors would like to thank to Engineering Faculty, and

35000
Industrial Engineering Department, Universitas Andalas for
30000 providing publication grant 2018.
25000
Demand

20000 CONCLUSIONS
15000
10000 Based on the analysis, the conclusions of this research are in 2016
by using EOQ method, the optimal order quantity is 32,073 ton
5000
per order, and the frequency is 9 times in a year with total cost Rp
0
4,757,673,813.48, and in 2017, the optimal order quantity is
25 27 29 31 33 35 37 39 41 43 45 47
34,856 ton per order and the frequency is 9 times in a year with
Period total cost Rp. 9,694,805,608.36. The Inventory cost that is spent
Figure 4. Demand Forecasting of Gypsum in PT Semen Padang by the company is not optimal, because the company still can
minimize the total inventory cost if they use the EOQ method.

DOI: 10.25077/josi.v18.n2.p116-124.2019 Fithri et al. 123

 FITHRI ET AL. / JURNAL OPTIMASI SISTEM INDUSTRI - VOL. 18 NO. 2 (2019) 116-124

REFERENCES [12] S. J. Taylor and B. Letham, “Forecasting at Scale,” The

American Statistician, vol. 72(1), pp. 37-45, 2018.
https://doi.org/10.7287/peerj.preprints.3190.
[1] E. M. B. Aske and S. Skogestad, “Consistent inventory
[13] C. C. Defee, B. Williams, W. S. Randall, and R. Thomas,
control,” Industrial & Engineering Chemical Research, vol.
“An inventory of theory in logistics and SCM research,”
48, no. 24, pp. 10892–10902, 2009.
The International Journal of Logistics Management, vol.
https://doi.org/10.1021/ie801603j.
21(3), pp. 404-489, 2010.
[2] S. Ziukov, “A literature review on models of inventory
https://doi.org/10.1108/09574091011089817.
management under uncertainty,” Business System and
[14] S. E. Sampson and M. Spring, “Customer Roles in Service
Economics, vol. 5(1), pp. 26-35, 2015.
Supply Chains and Opportunities for Innovation,” Journal
https://doi.org/10.13165/VSE-15-5-1-03.
of Supply Chain Management, vol. 48(4), pp. 30-50, 2012.
[3] D. O. N.G. Hai, Z. H. E. N. G. Hao and L. Y. Ping, “Model
https://doi.org/10.1111/j.1745-493X.2012.03282.x.
predictive control for inventory management in supply
[15] E. Van Der Laan, M. Salomon, R. Dekker, and L. Van
chain planning,” Procedia Engineering, vol. 15, pp. 1154-
Wassenhove, “Inventory control in hybrid systems with
1159, 2011. https://doi.org/10.1016/j.proeng.2011.08.213.
remanufacturing,” Management Science, vol. 45(5), pp.
[4] R. Chaib, M. L. Bouanaka, A. Bellaouar, M. Benidir, and I.
733-747, 1999. https://doi.org/10.1287/mnsc.45.5.733.
Verzea, “Determination of the efficiency of a means of
[16] J. S. Morris and R. J. Tersine, “A simulation comparison of
production application in a chain of production,”
process and cellular layouts in a dual resource constrained
Management of Technological Changes, vol. 1, pp. 605–
environment,” Computers & Industrial Engineering, vol.
608, 2009.
26(4), pp. 733-741, 1994. https://doi.org/10.1016/0360-
[5] A. Bacchetti, R. Plebani, N. Saccani, and A. Syntetos,
8352(94)90008-6.
“Spare Parts Classification and Inventory Management: a
[17] F. Lestari, Ulfah, Suherman, B. Azwar, and P. Fithri,
Case Study,” Salford Business School Working Papers
“Combining ABC and VED analysis for managing
Series, vol. 408, 2010.
medicine inventory (case study at community development
[6] G. A. Keskin, S. I. Omurca, N. Aydin, and E. Ekinci, “A
elderly in Indonesia),” International Journal on Advanced
comparative study of production-inventory model for
Science Engineering Information and Technology, vol.
determining effective production quantity and safety stock
9(3), 2019. https://doi.org/10.18517/ijaseit.9.3.7153.
level,” Applied Mathematical Modelling, vol. 39(20), pp.
[18] S. Assauri, Manajemen Produksi dan Operasi. Jakarta:
6359-6374.2015.
Lembaga Penerbit Fakultas Ekonomi Universitas
https://doi.org/10.1016/j.apm.2015.01.037.
Indonesia, 2008.
[7] D. K. Chester, A. M. Duncan, and C. J. L. Dibben, “The
importance of religion in shaping volcanic risk perception
in Italy, with special reference to Vesuvius and Etna,” NOMENCLATURE
Journal of Volcanology Geothermal Research., vol. 172(3–
4), pp. 216–228, 2008. D = Demand
https://doi.org/10.1016/j.jvolgeores.2007.12.009. Q = Lot size
[8] M. Godichaud and L. Amodeo, “Economic order quantity A = Ordering cost
for multistage disassembly systems,” International Journal
H = the holding cost to the purchase price/unit
of Production Economics, vol. 199, pp. 16–25, 2018.
https://doi.org/10.1016/j.ijpe.2018.02.008. C = the price per unit of material
[9] R. Kumar, “Economic Order Quantity (EOQ) Model,” B = reorder point
Global Journal of Finance and Economic Management, vol. ac = ce = interval between orders
5, no. 1, pp. 2249–3158, 2016. ab = cd = ef = lead time
[10] L. B. Schwarz, “The economic order-quantity (EOQ) m = Number of average inventory
model,” International Series in Operations Research and x = The number of demand
Management Science, vol. 115, 2008, pp. 135–154.
𝑥̅ = The average of demand
https://doi.org/10.1007/978-0-387-73699-0_8.
[11] K. C. Land, “Forecasting,” in International Encyclopedia of N = The number of the demand period
the Social & Behavioral Sciences: Second Edition, 2015.
https://doi.org/10.1016/B978-0-08-097086-8.10544-6.

124 Fithri et al. DOI: 10.25077/josi.v18.n2.p116-124.2019