Survey on Declining Curves of Unconventional Wells and Correlation with Key Determinants
2 likes118 views
The document discusses the analysis of decline curves in unconventional shale gas wells, focusing on the Marcellus shale as a case study. It highlights key production factors such as reservoir and fracture parameters, adsorption/desorption mechanisms, and the importance of numerical simulations for predicting gas recovery. The findings indicate the significant influence of horizontal well length and fracture characteristics on production rates and economic viability in shale gas extraction.
Survey on Declining Curves of Unconventional Wells and Correlation with Key Determinants
- 1. Survey on Declining Curves of Unconventional Wells and Correlation with Key Determinants Presented by: Borhen Addine AFLI Salman Deumah Rafaa Saadouli October 2020 CUPB: China University of Petroleum Beijing
- 2. Plan Introduction Scale of media porous in shale gas reservoirs Adsorption and desorption mechanism Flow mechanisms Numerical simulatoion and sensitivity analysis Decline curves analysis: ex. Marcellus shale Conclusion
- 3. Introduction • Shale gas is a natural gas resources. It is an unconventional resources produced from shale formations. • shale are sedimentary rocks characterized by low porosity and low permeability (matrix). • Flow mechanisms are related to the pores size, pressure and velocity. Darcy law is not always applicable. • The amount of gas in place is the sum of the free gas stored in the matrix and/or fracture porosity, and the adsorbed gas which is stored (adsorbed) on the surface of the organic or mineral material. • The most techniques used in shale gas production and for unconventional reservoirs are horizontal wells and multistage hydraulic fracturing. • The gas production will depend on the fracture parameters mainly and the reservoir parameters. • Numerical simulation has a big advantage on the prediction of the reservoir performance and the identification of the influence of each parameter. • The use of Decline curve analysis provide an important tool for the forecast of the production if the necessary assumption are taken into account. 1 20
- 4. Scale of media porous in shale gas reservoirs • The pores size has a significant function in controlling the flow in shale gas reservoirs. • Conventional reservoirs are characterized by pores size that range from 0.1 to 100 μm, for tight sandstone reservoir pore size range from 2 to 0.03 μm. While for shale reservoirs pores size range 1 to 200 nm and main pores size range from 2 to 50 nm (0.002 to 0.05 μm). • The size of hydrocarbon molecules, asphaltenes, paraffins and methane range from 0.01 μm to 0.00038 μm. • For this shale characteristics, the interaction between hydrocarbon molecules will affect the flow mechanism, as we have more interaction between molecule as we will have less amount of gas that can be extracted. This is mainly due to the size of the nanopores especially when we have high pressure. Example the slip flow • The interaction between hydrocarbon molecules is defined by the mean free path (how fare the molecule will travel before heating another molecule). • This effect is present in conventional reservoirs but due to the large pores size, the interaction between molecules has not a significant impact. 2 20
- 5. Adsorption/desorption mechanisms • Adsorption is a physic mechanism by which hydrocarbon molecules adhere to the surface of the organic or mineral material. • The desorption mechanism is the mechanism by which the liberation of gas take place with the decrease in the reservoir pressure. • They are important mechanisms that control the storage and production capacity in shale gas reservoirs. • Depend on the reservoir pressure, the adsorption could be monolayer or multilayer. Hence many authors try to explain this mechanism using adsorption isotherm curves. 3 20
- 6. Adsorption/desorption Adsorption isotherm: It is a curve that shows the amount of the adsorbed gas or the adsorbate by the solid or adsorbent (solid itself) as function of the relative pressure at constant temperature. Three main type of adsorption isotherm are: ▪ FreundlichAdsorption Isotherm: the first established adsorption isotherm. m x =K P1/n ▪ Langmuir Adsorption Isotherm: most used isotherm to quantify the adsorption mechanism at low reservoir pressure. c g = l V ∗ 𝑃 (P+Pl) ➢ Both model give satisfactory result at low reservoir pressure, and fail to quantify the amount of adsorbed gas at high pressure. ➢ Langmuir's model was a theoretical construct, while the Freundlich isotherm is empirical. In the Langmuir model, we assumed that there is only a monomolecular layer on the surface. This means that there is no stacking of adsorbed molecules Where, x :mass of the adsorbate; m: mass of the adsorbent; K and n are constant (depend on adosorbent and adsorbate characteristics); P is the reservoir pressure 4 20
- 7. Adsorption/desorption mechanisms ▪ Multilayer adsorption the BET equation: BET : Stephen Brunauer; Paul Emmett; Edward Teller ➢ Most widely used isotherm dealing with multilayer adsorption is BET Isotherm 5 20
- 8. Flow mechanism • The flow mechanism in shale gas reservoir is characterized with the Knudsen number (as Reynolds number). Knudsen number, Kn, is the ratio of mean free path, λ, to pore diameter, d, and can be used to identify different flow regimes in the porous media. • Slip flow: gas transport is dominated by collision between gas molecules • Knudsen diffusion (transition flow and free-molecule flow) for Kn larger than 0.1: gas transport is dominated by collision between gas molecules and the pore wall. Knudsen number (Kn) Flow regime Kn< 10-3 Continuum/darcy flow (no-slip flow) 10-3-10-1 Slip flow 10-1-101 Transition flow 101-∞ Free-molecule flow 6 20
- 9. Numerical simulation ▪ The numerical simulation model is used to predict the behavior of the reservoir taking into account the main influencing parameters. ▪ The use of enough data enhance the quality of the model and gives more confident results in term of forecast of production and decline production rate. ▪ The creation of the model passes through the experiments design (what are the important parameters) and the model function determination. ▪ Response surface methodology RSM is used to preform the creation of the model using Latin hypercube design (experiments design) and Genetic programing for the determination of the model function. ▪ After establishing the adequate model a sensitivity analysis study is performed to identify the importance of each parameter. Parameters are divided into two classes: reservoir parameters (matrix porosity and permeability, natural fracture porosity, etc.) and fracture parameters (fracture porosity and conductivity, spacing, etc.) ▪ Sensitivity analysis is performed with the establishment of a base case for the model. Each parameter have a certain value which determine its base case. All results in terms of cumulative production and decline gas production rate are compared with the result of the base case model. 7 20
- 15. Numerical simulation HF half length (23.04%). 13 20 For the first year of production, the gas production is mainly controlled by the fracture parameters (76%). HF spacing, HF conductivity and HF height contribute by around 66%. For ten years of production, the gas production is mainly controlled by the fracture parameters (85%) As we decrease the HF spacing, we create new porosity and we enhance the permeability. An increase of the effect of HF spacing is due to the decrease of the effect the matrix porosity and permeability. For twenty years of production, fracture parameters participate by 59% to the total production.
- 16. Decline Curve Analysis: Marcellus Plays One of the richest gas field in North America. Extent primarily in Pennsylvania, West Virginia, New York, and Ohio. Devonian organic- rich shale (middle Devonian) that has an area of 18 106 acres (72,843 km2) OGIP= 1500 TCF Estimated reserves= 280 TCF Has a big impact over the Pennsylvania economic. 14 20
- 17. Decline Curve Analysis: Marcellus Plays Started in production in 2008, more than 5,400 Marcellus wells were on line in June 2014 and more than 1,200 wells are being added each year. data of 3845 horizontal wells is used to perform the decline curve analysis in order to forecast the Estimated Ultimate Recovery EUR of the field. No detailed information is available concerning the exact procedure used in the decline curve analysis. From the available data we can conclude that the Hyperbolic decline curve equation is used to perform the determination of the EUR distribution. The average Value of the hyperbolic b-factor is around 0.52 which is in favor of the use of the hyperbolic decline model. The decline curve analysis is applied each year of production which gives the possibility to determine the average decline rate. The calculation of the correlation coefficient gives the possibility to link the different parameters. 15 20
- 18. Decline Curve Analysis: Marcellus Plays Year of production number of wells Average initial production MCF/month average horizontal length ft average initial decline rate % Hyperbolic factor b EUR Bcfe 2008 28 43263 2280 43,87 0,65 2032,69 2009 174 71768 2890 43,21 0,55 3327,25 2010 532 121248 3800 48,52 0,65 4873,21 2011 659 121339 4100 49,02 0,66 4531,52 2012 1254 129302 4500 48,38 0,51 4201,77 2013 1198 172787 4751 46,22 0,39 5397,4 Somme 3845 Average 4324 0,52 Correlation Average initial production MCF/month EUR (BCFe) Number of wells 0,90 0,76 Average horizontal length (ft) 0,97 0,91 Average initial production (MCF/month) - 0,95 Average horizontal length has more effect on the EUR than the number of wells. Average horizontal length shows a very important correlation with the average initial production. (Homogeneity). Average initial production has a good correlation with the EUR. As we produce more at the beginning we will have good recovery. ( effect 16 20 number of wells, Hz length)
- 19. Decline Curve Analysis: Marcellus Plays Horizontal length ft Aveg. gas flow test MCFPD Avg. Max gas production MCF/month Hyperbolic b factor EUR per well Average 3956 6033 126503 0,52 0,02 Standard deviation 790,8 3067,2 64891,7 0,22 0,33 ➢ The average gas flow test and the average initial gas production shown a good correlation with the horizontal length. ➢ Standard deviation factor shows the important distribution of the EUR. This distribution is mainly due to the horizontal well length which shows a standard deviation of 791 ft. 17 20
- 20. Conclusion • Unconventional resources has a big economic impact, hence it represents an important tool to ameliorate the economical situation of any state. • Horizontal well and hydraulic fracture are the main technological tool that we can use to achieve the economic production. • Fracture parameters are the main parameters that control the production in shale gas reservoirs. • Gas desorption plays an important role in the gas production. Significant effect of the gas desorption is achieved when we use different Langmuir parameters. Case 1 Case 2 Case 3 No gas desorption • Decline curves analysis shown that the horizontal length is an important parameter that control the EUR Langmuir Pressure (psi) 400 1000 1500 N/A Langùuir volume (SCF/ton) 40 140 220 N/A 20 yr. cumulative production (MMscf) 2504 2646 2775 2449 18 20
- 21. References • A multiscale-multiphase simulation model for the evaluation of shale gas recovery coupled the effect of water flowback. Peng, Cao a, Liu, Jishan and Leong, Yee-Kwong. 2017. 2017. • A Note on Rock Compressibility. Onur, M. 2017. s.l. : Spring, 2017. •Adsorption of gases in multimolecular layers. Brunauer, S et al. s.l. : JAm Chem Soc. Vol. 60, pp. 309-319. •Applicability of the Forchheimer Equation for Non-Darcy Flow in Porous Media. H. Huang, and J.Ayoub. 2006. s.l. : SPE 102715, September 2006. • Dynamic Data Analysis. KAPPA. 1988-2017. 1988-2017, Vol. v5.12.01. • Dynamics of Fluids in Porous Media. BEAR J. 1972. Dover, New York : s.n., 1972. •Genetic Programming-Computers using Natural Selection" to generate programs. William, B. LangdonAdil Qureshi. s.l. : Dep of Computer science, University College London. •Guo, C Xu J Wu K Wei M Liu S. 2015. Study on gas flow through nano pores of shale gas reservoirs. 2015. pp. 107-117. 19 20
- 22. References • htt1 http://www.brad.ac.uk/staff/vtoropov/burgeon/thesis_luis/chapter3.pdf. •Javadpour. 2009. Nanopores and apparent permeability of gas flow in Mudrocks (shales and siltstone). J. Can. Petroleum Technology 48 (2009) 16-21. 2009. •Jun Y., Hai S.,Dong-yan F., Chen-chen W., Zhi-xue S. 2013. Numerical simulation of gas transport mechanisms in tight shale gas reservoirs. s.l. : Petroleum Science, 2013. Vols. Volume 10, Issue 4, pp 528-537. • K.S. Lee, and T.H. Kim,. 2016. Integrative Understanding of Shale Gas Reservoirs,. s.l. : SpringerBriefs inApplied Sciences and Technology, 2016. •Loucks, R Reed R Ruppel S. and Hammes. Spectrum of pore types and networks in mudrocks and a descriptive classification for matrix-related mudrocks pores. s.l. : The American Association of Petroleum Geologists Bulletin, pp. v. 96, No. 6, pp. 1071–1098. •Morphology, genesis and distribution of nano-scale pores in siliceous mudstones of the Mississippian Barnett Shale. J. Sed. Res. Loucks, R.G., et al. 2009. 2009. 20 20
- 23. Thank you for your attention