Introduction

Habiganj gas field, Bangladesh, was the study area to carry out this research. This gas field is located at Madhabpur Upazila, Habiganj District, about 100 kilo meter (km) away to the direction of northeastern from Dhaka (capital city of Bangladesh). This gas field lies about 32 km northeast of Titas gas field, the largest reserved gas field in Bangladesh. This gas field was discovered by shell Oil Company in 1963 and has been one of the major gas producers in Bangladesh (Imam 2013; Bangladesh Gas Fields Company Limited [BGFCL] 2014). Figure 1 shows the location of Habiganj gas field in Habiganj District in Bangladesh. It is the third-largest gas producer in the country and operated by Bangladesh Gas Field Company Ltd (BGFCL), a subsidiary of Bangladesh Oil, Gas and Mineral Corporation (known as Petrobangla) (Imam 2005).

Fig. 1
figure 1

Source: Wikipedia

Location of Habiganj gas field, Bangladesh

Full size image

There are two gas zones, upper gas sand (UGS) and lower gas sand (LGS). These two zones are sandstone formations. The UGS is the primary gas reservoir of the gas field and lies at a depth of 1320 m below the surface. It has a maximum gross pay of 230 m thick. The gas sand is medium to fine grained, well sorted, clean and unconsolidated. It has an average porosity of 30 % and average permeability in the range of 2–4 darcy (Imam 2013). Figure 2 represents the cross section of the subsurface of Habiganj gas field, Bangladesh.

Fig. 2
figure 2

Cross-sectional view of the subsurface of Habiganj gas field, Bangladesh (Imam 2013)

Full size image

According to Haq and Gomes (2001), Intercomp-Kanata Management Limited (IKM) Reservoir Engineering Report (1990) stated that the recovery from the UGS in the Habiganj gas field is dominated by the water drive mechanism and the aquifer to be ten times extensive than the reservoir. Hartmann and Beaumont (2016) says that water drive mechanism exists only where the aquifer is of equal or better quality and has a much larger volume than the reservoir (about 10 times).

Rate transient analysis

Rate transient analysis (RTA) is the science of analyzing production data (both rates and flowing pressures). This method is an important tool to estimate reserve of oil and/or gas of a reservoir. Reserve estimation and development planning are the key tasks of petroleum engineers by the use of historical production (reservoir fluid production rate histories and cumulative production). Both of these fall within the domain of a quantitative production data analysis (PDA) (Help Manual 2010; Clarkson et al. 2012; Clarkson 2013).

RTA can also be defined as a modern decline curve analysis (DCA) method (Mishra 2014). DCA method is one of the oldest and most often used tools of the petroleum engineers. This is a forecasting technique which predicts by history matching of rate-time data on an appropriate typecurve. What direction to take, what typecurve(s) to choose and where the rate-time data should fit are decided based on basic reservoir engineering concepts and knowledge (Fetkovich 1980; Fetkovich et al. 1987). So DCA is not based on applying a purely empirical equation to be analyzed with statistical approaches which often leads to unrealistic and unreliable forecast and reserve estimate (Fetkovich et al. 1996).

RTA also gives convincing estimates of reservoir parameters with available low-frequency (weekly or monthly) production data. Once reservoir characteristics are determined using RTA, a reservoir model is then constructed to forecast future production scenarios (Mishra 2014; Mireault and Dean 2007–2008).

Typecurve analysis

According to Ley and Samaniego (1981), a considerable amount of information concerning well test analysis has been in the literature over the last several decades. Typecurve analysis consists of finding a typecurve that matches the actual response of the well and the reservoir during the test. Then the reservoir and well parameters, such as permeability and skin, can be calculated from the dimensionless parameters defining that typecurve (Gringarten 1987).

Gas initial in place and expected ultimate recovery

Gas initial in place (GIIP) refers to the total amount of gas present initially in the underground of a gas field. Part of the GIIP in an explored gas field can be recovered. Generally, the recovery of gas from the GIIP in a typical gas field ranges from as low as 60 % to as high as 90 % (Imam 2013).

Expected ultimate recovery (EUR) of a petroleum source is the sum of the proven reserve at a specific time and the cumulative production up to that time. Proven reserve refers to the quantity of gas in a gas reservoir which can be estimated with reasonable certainty (high degree of confidence) to be commercially recoverable from known reservoir under the present economic and operating conditions (Imam 2013; Morehouse 1997).

Permeability and skin

Permeability (k) in a reservoir rock is its capacity to transport fluids through a system of interconnected pores. Reservoir permeability is a random-valued property of the formation (Zolotukhin and Ursin 2000; Jensen et al. 1987).

The pressure drop in a well per unit rate of flow is controlled by the resistance of the formation, the viscosity of the fluid, and the additional resistance concentrated around the wellbore due to drilling, completion and production practices. The pressure drop caused by this additional resistance is defined as the skin effect, denoted by the symbol S. The reservoir damage occurs because of this skin effect. Reservoir damage is resulted from well drilling and completion (S d), pseudo-skin factor resulted from reservoir open level (S PT) and pseudo-skin factor resulted from perforation (S PF) and several other skin factors (Everdingen 1953; Jianchun et al. 2014).

A two-region reservoir model (Fig. 3) represents a damaged well. In this model, the altered zone around the wellbore is assumed to have uniform permeability k s out to a radius r s, beyond which the formation permeability, k, is unaltered (Altered zone and skin effect 2016). Everdingen found that better agreement between theory and well performance is possible, if the permeability reduction near the wellbore region is assumed due to drilling, completion and production practices (Hawkins Jr. 1956).

Fig. 3
figure 3

Two-region reservoir model (Altered zone and skin effect 2016)

Full size image

Materials and methods

Data (reservoir properties, fluid properties, properties of Well-07 and Well-10, and production data of Well-07 and Well-10 for the year of 2007) were collected from BGFCL, a subsidiary of Petrobangla for this study. By using software FEKETE, F.A.S.T.RTA™, methods such as Blasingame typecurve analysis, Agarwal–Gardner typecurve analysis and normalized pressure integral (NPI) typecurve analysis were followed. After entering data into the software, data points in the graph were seemed to be very scattered. It was essential to be filtered. To minimize error and to obtain a good match between typecurves and data plot, data points in the graph of the software were filtered to clean up “noise” and to reduce data which were scattered.

The major assumptions in this work were as follows:

  1. 1

    The reservoir is assumed to be homogeneous.

  2. 2

    The reservoir is circular of radius r e.

  3. 3

    The fluid is of constant viscosity, μ.

  4. 4

    Radial flow of fluids.

Required data for the study

The required reservoir, fluid and well properties are numerated below. A table concerning the report of daily flowing well head pressure (FWHP) and daily production of gas and water for the year of 2007 are presented in Appendices 1 and 2 section in this paper.

Reservoir properties

  • Producing zone: upper gas sand

  • Porosity: 30 %

  • Reservoir temp: 178 degree F

Fluid properties

  • Fluid type: gas

  • Fluid used for static calculation: gas

Properties of Well-07

  • Well type: vertical

  • Perforation interval: 4347–4600 ft

  • Midpoint perforation: 4473.5 ft

  • Tubing size: 4.5 in

  • R w: 0.350 ft

  • Casing (ID): 7 in

  • Static Well head temp: 75 degree F (avg)

  • Flowing Well head temperature: 95 degree F (avg)

Properties of Well-10

  • Well type: vertical

  • Perforation interval: 4373–4649 ft

  • Midpoint perforation: 4511 ft

  • Tubing size: 4.5 in

  • R w: 0.350 ft

  • Porosity: 30 %

  • Casing (ID): 7 in

  • Static Well head temp: 75 degree F (avg)

  • Flowing Well head temperature: 95 degree F (avg)

Blasingame typecurve analysis

At Blasingame method (Help Manual 2010), the normalized rate was plotted against material-balance pseudo-time on a log–log scale of the same size as the typecurves, which are referred as the “data plot”.

$${\text{For}}\,{\text{data}}\,{\text{plot}},{\text{normalized}}\,{\text{rate}},\frac{q}{{\Delta {\text{p}}_{\text{p}} }} = \frac{q}{{(p_{i} - p_{\text{pwf}} )}}$$
$${\text{Material}}\,{\text{balance}}\,{\text{time}}\,t_{c} = \frac{Q}{q}$$

Application of this concept to oil would be very straight forward. To gas, it is more complex because of varying PVT properties of gas. Accordingly simple concept \(\mathop t\nolimits_{c} = \frac{{\mathop G\nolimits_{g} }}{{\mathop q\nolimits_{g} }}\) must be defined in terms of pseudo-time,

$$\mathop t\nolimits_{\text{ca}} = \left\{ {\frac{{(\mu_{\text{g}} c_{\text{g}} )_{j} }}{{q_{\text{g}} }}} \right\}\int_{o}^{t} {\left\{ {\frac{{q_{\text{g}} }}{{\mu_{\text{gav}} c_{\text{gav}} }}} \right\}} {\text{d}}t$$

Blasingame et al. established his typecurves using dimensionless rate \((q_{\text{Dd}} )\) against dimensionless time \(\left( {t_{\text{Dd}} } \right)\) on a log–log scale.

$$\begin{aligned} q_{\text{Dd}} & = q_{\text{d}} \ln \left( {r_{\text{eD}} - 0.5} \right),r_{\text{eD}} = \frac{{r_{\text{e}} }}{{r_{\text{w}} }} \\ t_{\text{Dd}} & = \frac{{\mathop {2t}\nolimits_{d} }}{{\left\{ {(\ln r_{\text{eD}} ) - 0.5)(r_{\text{eD}}^{2} - 1)\} } \right.}} \\ \end{aligned}$$

Agarwal–Gardner typecurve analysis

At Agarwal–Gardner method (Help Manual 2010), the procedure is almost similar to the Blasingame with a few dissimilarities. As opposed to Blasingame, here for data plot,

$${\text{Normalized}}\,{\text{rate}},\frac{q}{{\Delta p_{\text{p}} }} = \frac{q}{{(p_{\text{i}} - p_{\text{pwf}} )}}$$
$${\text{Material}}\,{\text{balance}}\,{\text{pseudo}}\,{\text{time,}}\,\mathop t\nolimits_{\text{ca}} = \left\{ {\frac{{(\mu_{\text{g}} c_{\text{g}} )_{j} }}{{q_{\text{g}} }}} \right\}\int_{o}^{t} {\left\{ {\frac{{q_{\text{g}} }}{{\mu_{\text{gav}} c_{\text{gav}} }}} \right\}} {\text{d}}t$$

Agarwal et al. established his typecurves using dimensionless rate (\(q_{\text{Dd}}\)) against dimensionless time \((t_{\text{Dd}} )\) on a log–log scale as,

$$\begin{aligned} q_{\text{D}} & = \frac{1}{{p_{\text{D}} }} = 141.2\frac{qB\mu }{{\left\{ {kh\left( {p_{\text{i}} - p_{\text{wf}} } \right)} \right\}}} \\ t_{\text{DA}} & = \frac{0.00633kt}{{\varphi \mu c_{t} A}} \\ \end{aligned}$$

Normalized pressure integral typecurve analysis

In case of NPI (Help Manual 2010), the normalized pressure replacing normalized rate was plotted against material-balance pseudo-time on a log–log scale of the same size as the typecurves, which are referred as the “data plot”.

$${\text{Normalized}}\,{\text{pressure}},\frac{{\Delta p_{\text{p}} }}{q} = \frac{{(p_{\text{pi}} - p_{\text{wf}} )}}{q}$$
$${\text{Material}}\,{\text{balance}}\,{\text{pseudo}}\,{\text{time,}}\,t_{\text{ca}} = \left\{ {\frac{{(\mu_{\text{g}} c_{\text{g}} )_{j} }}{{q_{\text{g}} }}} \right\}\int_{o}^{t} {\left\{ {\frac{{q_{\text{g}} }}{{\mu_{\text{gav}} c_{\text{gav}} }}} \right\}} {\text{d}}t$$

NPI typecurves were developed by dimensionless pressure (\(p_{\text{d}}\)) against dimensionless time (\(t_{\text{d}}\)) on a log–log scale.

$$p_{\text{D}} = \frac{1}{{q_{\text{D}} }} = \frac{{\left\{ {kh\left( {p_{\text{i}} - p_{\text{wf}} } \right)} \right\}}}{141.2qB\mu }$$
$$t_{\text{DA}} = \frac{0.00633kt}{{\varphi \mu c_{\text{t}} A}}$$

Analysis procedure

After entering data in the software FEKETE, F.A.S.T.RTA™, instruction of the Help Manual was followed. According to the Help Manual, typecurve analysis is generally done by selecting a match point and reading its coordinates off the data plot (\(\frac{q}{p}\) and \(\mathop t\nolimits_{\text{ca}}\)) and off the typecurve plot (\(q_{\text{Dd}}\) and \(t_{\text{Dd}}\)) match. At the same time, the stem value \(\frac{{\mathop r\nolimits_{\text{e}} }}{{\mathop r\nolimits_{\text{w}} }}\) of the matching curve is noted. In this study, data plot was moved over the typecurve plot to get the optimum match, while the axis of the two plots was kept parallel. Several positions of the several typecurves were tried and typecurve that fitted the data plot best was selected. From a curve match, the following reservoir parameters were found from the output of the software: GIIP, EUR, permeability and skin.

Results and discussions

Using three methods, graph of the match between data plot and individual typecurves are presented in Figs. 4, 5, 6 for Well-07 and Figs. 7, 8, 9 for Well-10, respectively. To get the optimum value of the results, three methods were averaged.

Fig. 4
figure 4

Match between data plot for Well-07 and Agarwal–Gardner typecurve plot

Full size image
Fig. 5
figure 5

Match between data plot for Well-07 and Blasingame typecurve plot

Full size image
Fig. 6
figure 6

Match between data plot for Well-07 and NPI typecurve plot

Full size image
Fig. 7
figure 7

Match between data plot for Well-10 and Agarwal–Gardner typecurve plot

Full size image
Fig. 8
figure 8

Match between data plot for Well-10 and Blasingame typecurve plot

Full size image
Fig. 9
figure 9

Match between data plot for Well-10 and NPI typecurve plot

Full size image

GIIP and EUR

GIIP of Well-07 is shown in Figs. 4, 5, 6 and of Well-10 is shown in Figs. 7, 8, 9, respectively. Table 1 shows the findings of GIIP values of Well-07 to be 435.082 Bcf and of Well-10 is 475.242 Bcf, respectively, obtained from FEKETE, F.A.S.T.RTA™. It could be said that the difference of GIIP between these two wells is very small.

Table 1 GIIP of Well-07 and Well-10
Full size table

EUR of Well-07 is shown in Figs. 4, 5, 6 and of Well-10 is shown in Figs. 7, 8, 9, respectively. As shown in Table 2, the EUR value of Well-07 and Well-10 obtained from FEKETE, F.A.S.T.RTA™ was projected to be 304.558 Bcf and 332.67 Bcf, respectively. This result suggests that only this amount of reserve from GIIP of these two wells mentioned above is commercially recoverable by primary recovery.

Table 2 EUR of Well-07 and Well-10
Full size table

Skin and permeability

The value of permeability (k) in the wellbore region of the Well-07 is illustrated in Figs. 4, 5, 6 and of Well-10 in Figs. 7, 8, 9, respectively. According to Table 3, permeability was estimated to 3.0396 md and 2.7839 md surrounding the Well-07 and Well-10, respectively, at the end of the year of 2007.

Table 3 Permeability (k) surrounding Well-07 and Well-10
Full size table

The value of the skin effect (dimensionless parameter) is depicted in Figs. 4, 5, 6 for Well-07 and for Well-10 in Figs. 7, 8, 9, respectively. As shown in Table 4, reservoir was damaged due to skin in surrounding region of the Well-07 was valued by 7.017 and of Well-10 was 7.014, respectively.

Table 4 Skin effect for Well-07 and Well-10
Full size table

Comparison with previous studies

Imam (2013) stated that the average permeability (k) value of UGS of Habiganj gas field was found to be 2–4 darcy. But from this study, the permeability (k) value of UGS of Habiganj gas field was obtained to 3.0396 md and 2.7839 md surrounding Well-07 and Well-10, respectively. Additionally, the dimensionless value of skin factor (s) was estimated surrounding the Well-07 and Well-10 to 7.017 and 7.014, respectively. So, permeability near these two wellbore regions was substantially reduced as a result of the highly valued skin effect as recommended by Hawkins Jr. (1956). This reduced permeability causes an additional pressure drop across the altered zone.

Imam (2013)also mentioned that Petrobangla published a revised estimate of the Habiganj gas field based on RPS energy consultant showing estimated GIIP value of UGS is 3.68 Tcf and EUR of 2.63 Tcf. This study found total estimated value of GIIP of these two wells to 910.324 Bcf and EUR to 637.228 Bcf. So, it can be deduced that 637.228 Bcf of gas is recoverable by primary recovery from proven reserve of 2.63 Tcf from the reservoir of Habiganj gas field, Bangladesh.

Conclusions

For the last few decades RTA, i.e., modern DCA, using typecurve is one of the developed and most used methods for analyzing petroleum reservoir. Averaging from three methods rate transient analyzing procedure, this study shows the estimated GIIP value of Well-07 and Well-10 of Habiganj gas field to 435.082 Bcf and 475.242 Bcf, respectively, while the value of EUR of Well-07 and Well-10 was amounted to 304.558 Bcf and 332.67 Bcf, respectively. The study shows the positive skin factor caused by reservoir damage for Well-07 and Well-10 was 7.017 and 7.014, reducing the value of formation permeability from 2–4 darcy to 3.0396 md and 2.7839 md near wellbore region. This decreasing permeability due to formation damage retards the expected flow of fluids and, hence, the production.