SYSTEM PREDICTION PRODUCTION PT . VICO INDONESIA USING METHOD HOLT WINTERS

Problem that Taken in this study is the process of forecasting oil and gas production in accordance so that companies can know the prediction of the amount of oil and gas in the future. The method used to determine production prediction is Holt-Winters forecasting method. In testing the system will do the comparison of alpha, beta and gamma. Using the alpha value = 0.2, beta = 0.1 and gamma 0.5 to get better multiplicative forecast for oil and gas data. And to get the smaller error difference compared to the smaller alpha (α), beta (β) and gamma (γ) then the smaller the difference will be. The Multiplexative Spring Method and the Seasonal Additive Method are good enough for oil and gas production data


INTRODUCTION
PT. VICO Indonesia is a company engaged in the field of oil and gas.PT.VICO Indonesia has operated in Sanga-Sanga Production Sharing Contract (PSC), located in Kutai Basin of East Kalimantan and covers an area of approximately 1,700 square kilometers, for more than 40 years.This has resulted in more than 12.6 TCF of gas and 0.4 billion barrels of liquid from production fields in Badak, Mutiara, Semberah, Nilam, Pamaguan, Lampake and Berau.In drilling not only oil and gas are drawn from the bowels of the earth but there is also water and condensate.Then there will be a filtering process that will separate water, condensate, oil and gas.
Forecasting is an activity to predict what will happen in the future.There are many types of forecasting.For example smoothing method, Jenkins Box method and trend projection method with regression.However, since the data is seasonal the most suitable method is the exponential smoothing method of Holt-Winters which sees in terms of seasonality in a data.
If this final project aims to implement Holt-Winters method to predict oil and gas production in PT.VICO Indonesia is based in Semberah, East Kalimantan using data from 2011, 2012, 2013, 2014 and 2015.With this implementation in order to provide an overview of production in PT.VICO Indonesia.

ANALYSIS AND DESIGN SYSTEM
To cope Increasing or decreasing the amount of oil and gas currently required a forecasting application that can help in improving the effectiveness of company performance.In building a forecasting application, it needs accurate data, validity, data adequacy and can describe time series.Therefore, it is necessary to perform several stages of analysis such as: a. Collecting data oil and gas From 2011 to 2015.b.Enter the data amount of oil and gas into the application system for forecasting.The following is data on the amount of oil and gas at PT.VICO Indonesia in the samberah plant.From data oil and gas above can be determined value alfa ( ) sebesar 0.2 , nilai beta ( ) 0.1 and gamma ( ) 0.5.Forecasting is done using the method Holt Winters With a seasonal multiplication factor (Multiplicative Seasonal) , Because the existing data are fluctuating always decrease and increase in every month and method of smoothing Exponential Holt-Winters With Seasonal Addition Method (Additive Seasonal Method) Which is used for constant seasonal variations.

Flowchart
Part plot system (flowchart) is a part of Which explains in detail the steps of the system process.To start this system the admin is required to login first then input the data of oil or gas, then specify the value alfa (α), Beta (β) and Gamma (γ), Then do the calculation using Holt -Winter's, After that calculate the level of forecasting errors using MSE and SSE, then out the calculation of forecasting the amount of oil and gas.
Using the same formula used to calculate until December of 2016.
Here is an example calculation to get the next value: Using the same method used to calculate the next period until December of 2016.
Here is a comparison table of real data for 2013 with forecasting results for 2013.Using the same formula is used to calculate the next period until December of 2016.
Here is an example calculation to get the next value.Using the same method used to calculate the next period until December of 2016.
Here is a comparison table of real data for 2013 forecasting results with 2013.

RESULT AND DISCUSSION
From the data obtained can be concluded that use Holt-Winters Multiplikative seasonal methot More accurate in predicting oil and gas production data year 2016 at PT.VICO Indonesia plan samberah.This is a forecasting chart using Alfa = 1 beta = 1 gamma = 1:

4
Based on the above formula, the calculation of the value of MSE oil from the results of 2013 is: Based on the above formula, the calculation of SSE oil values from 2013 data are: Calculation Holt Winter's Using Seasonal Additions (Additive Seasonal Method) Of the oil data.1) How to Find the Initial Value and in Additive Same as Multiplikative.2) Example calculation to find the initial value ( ) Period 25 is as follows: = X − S .

Table 2 .
1 data Oil and gas PT.VICO Indonesia monthly.

.2 Flowchart Program (Multiplicative Seasonal Method)
1 Flowchart system Forecasting for Oil and Gas Picture following describes the flow of forecasting by method Holt Winter's Using seasonal multiplication adjustment factors (Multiplicative Seasonal Method).

.3 Flowchart Program (Additive Seasonal Method)
Picture following describes the flow of forecasting by Method Additions Seasonal (Additive Seasonal Method).

3 Calculation Holt Winter's Using Seasonal Multiplication (Multiplikative Seasonal Method) Of the oil data. 1
) Here is a calculation forecasting Holt Winter's Using multiplication seasonal, To get the initial value of the period 2) Example calculation to find the initial value ( ) For period 25th from 2011 to 2012 data is as follows : Tabel 2.2 Oil and Gas Forecasting Results of 2013 Using Seasonal Multiplication Factor.
Tabel 4.6 Forecasting Oil and Gas Results 2013 Using Season Summation Factors