Simulation of underground storage / UM EL-Radhuma Formation-Ratawi field

The aim of this study is to investigate the feasibility of underground storage of gas in Um ElRadhuma formation /Ratawi field. This formation is an aquifer consisting of a high permeable dolomite beds overlain by impermeable anhydrite bed of Rus formation. Interactive petrophysics (IP), Petrel RE and Eclipse 100 softwares were used to conduct a well log interpretation, build a reservoir simulation model and predict the reservoir behavior during storage respectively. A black oil, three dimensional and two phase fluid model has been used. The results showed that the upper part of Um El-Radhuma formation is suitable for underground gas storage, because of the seal of its cap rock and capability of reserving gas in the reservoir. It was found that available volume for storage is 14.3 billion cubic feet with a structural closure of 45 m. the optimum injection rate has been calculated also. Introduction Natural gas is playing an important role in meeting the world’s energy demand because of the clean burning, huge energy that can be supplied by using it. According to the International Energy Agency, in the next 20 years, the average growth rate of global natural gas demand is expected to be over 1.8% per year [1]. Also, natural gas is known that it is mutable in its demand according to the year seasons, if it is cold or hot [2, 3]. Today in Iraq, large volumes of natural gas are not invested and lost by burning, because the main type of natural gas in Iraq is associated gas and needs high capitals of money to be invested. For example, more than1.5 MMMSCF/d were burned in 2016 which means 1.2 MMMSCF of dry gas, 7500 ton of liquefied gas and 1650 ton of benzene have been lost [4]. Natural gas storage is used to face the expected future peak demand, maintain the balance between supply and demand and to save the noninvested gas from burning and losing [3]. No.19 Journal of Petroleum Research & Studies (JPR&S)


Introduction
Natural gas is playing an important role in meeting the world's energy demand because of the clean burning, huge energy that can be supplied by using it. According to the International Energy Agency, in the next 20 years, the average growth rate of global natural gas demand is expected to be over 1.8% per year [1]. Also, natural gas is known that it is mutable in its demand according to the year seasons, if it is cold or hot [2,3].
Today in Iraq, large volumes of natural gas are not invested and lost by burning, because the main type of natural gas in Iraq is associated gas and needs high capitals of money to be invested. For example, more than1.5 MMMSCF/d were burned in 2016 which means 1.2 MMMSCF of dry gas, 7500 ton of liquefied gas and 1650 ton of benzene have been lost [4]. Natural gas storage is used to face the expected future peak demand, maintain the balance between supply and demand and to save the noninvested gas from burning and losing [3].

Methodology
During this study, two phase and 3D black oil reservoir simulation model was built to present the aquifer behavior during injection processes. Therefore, Petrel and Eclipse softwares were used to build and simulate the reservoir model. The five main stages that have been conducted during this study were:-1-Well logs interpretation: during this stage all the data of well logs have been used to find CPI for every well such as porosity, clay volume, water saturation. Well logs interpretation was done by using interactive petrophysics (IP) software. Wells that have been used during for this purpose were RT-3, 6, 7, 8, 9, 10 and 11. Also, fracture pressure has been calculated during this study by using modified Eaton method and interactive petrophysics (IP) software.

4-History matching:
This stage is very important to verify that the built geological model that was built matches the observed rates and pressure. This is used to prove that the model gives the reasonable prediction result for the future. Therefore, depending on DST tests of well RT-10 for several days, a history matching of the observed data of water production rate and bottom well pressure has been obtained as shown in figure (4) and figure (5) respectively.

5-Predict reservoir behavior during injection:
In order to predict reservoir behavior during injection, 9 gas injectors had been added to the model, in addition to wells RT-8 and RT-10.  x Injection without injection rate target but, constant BHP of 1650 (case_ D).
x Dividing the 10 years simulation period into 5 periods to find optimum injection rate of every period (case_ E). From figure (7), it can be noticed that injection rate target more than 51 MMSCF/D cannot be sustained. The injection rate continues decreasing after a periods depending on the value of injection rate. This is because of continuous injection leads to an increasing in the formation pressure figure (8) which leads to a decrease in the differential pressure and the last affects injection rate proportionally; therefore, the optimum injection rate of the 10 years injection periods without pressure relief is 51

Fig. (8) Average reservoir pressure, reservoir and surface cumulative gas injected of the cases A, B and C during 10 years injection period
The reservoir & surface cumulative injected gas volume and average formation pressure of the above three cases after 10 years of injection are described in table (1).

Table (1) Average reservoir pressure, surface and reservoir cumulative gas injected after 10 years of case A, B and C.
In order to get more gas injection rate and to determine the maximum cumulative gas injected, two cases have been conducted, the first is without injection rate target (case_ D) and the second is with several injection rate periods (case_ E).       From the comparison between figure (7) and figure (9), it can be noticed that discretization the injection period into several periods with several injection rate targets leads to an increase in the optimum injection rate, which leads to an increase in the cumulative gas injected figure (8 & 10).
Therefore, increasing the injection periods leads to an increase in optimum injection rate target.
Optimum injection rate of every period during case E are described in table (3) which is the injection rate target of every period as shown in figure (9).