##
Rezervuar Heterojenliğinin Su Öteleme Performansı Üzerindeki Etkileri

Rezervuar Heterojenliğinin Su Öteleme Performansı Üzerindeki Etkileri

##### Dosyalar

##### Tarih

1996

##### Yazarlar

Tarancı, Teoman

##### Süreli Yayın başlığı

##### Süreli Yayın ISSN

##### Cilt Başlığı

##### Yayınevi

Fen Bilimleri Enstitüsü

Institute of Science and Technology

Institute of Science and Technology

##### Özet

Bilindiği gibi, birincil üretim ile rezervuardaki petrolün, kayaç ve rezervuar akışkanlarının özelliklerine göre, % 15 ile 45 arasında üretilebilmektedir. Bu üretim rezervuardaki başka faktörlerinde etkisi ile % 5-10 civarına düşebilmektedir. Bunun sonucunda uygulanan üretim artırma yöntemleri ile kalan petrolün üretilmesi hedeflenmektedir. Rezervuarda üretim ile azalan rezervuar basıncını artırmak ve daha fazla petrolü öteleyerek üretebilmek amacı ile en yaygın ve ekonomik yöntem olan su enjeksiyonu yöntemi uygulanmaktadır. Bu çalışmada, petrol endüstrisinde çok yaygın olarak kullanılan ve yazılım kodu sağlanabilir Amerika Enerji Kurumu' nca geliştirilmiş olan BOAST II (Black Oil Applied Simulation Tool) yazılımı, heterojen sistemlerde su enjeksiyonu performans değerlendirmesi için kullanılmıştır. BOAST II bilindiği gibi gözenekli ortamda çok boyutlu ve çok fazlı akış sistemleri için geçerli olduğundan su enjeksiyon performansında olduğu kadar yeraltı sularının akışının modellenmesinde de kullanılması mümkündür. Rezervuar heterojenliği, öteleme performansını büyük çapta etkileyen en önemli faktörlerden biridir. Bundan dolayı, karmaşık akış koşullarında heterojenliğin etkilerini incelemenin en etken yolu sayısal modellemedir. Yapılan çalışmada ilk olarak, bir boyutlu sistemler için BOAŞT II, Buckley- Leverett analitik yaklaşımı ile test edilmiş ve elde edilen çözümlerin oldukça uygun olduğu görülmüştür. Çalışmada ağırlıklı olarak iki boyutlu alansal heterojenliğin öteleme performansı üzerindeki etkileri çalışılmıştır. Ayrıca dikey yönde de bazı örnekler göz önüne alınmış olup, hem korelasyonlu hem de korelasyonsuz dağılım gösteren log-normal geçirgenlikler gözönüne alınmış ve beş-nokta paterninde sistemi temsil etmesi beklenen geometrik ortalama geçirgenlik değerleri ile yapılan performans değerleri ile kıyaslanmıştır. Heterojenliklerin hangi durumlarda performans üzerinde etkili olabileceği gösterilmeye çalışılmıştır. Ayrıca kılcallığın su ötelemesi işlemlerinde gözenekli ortamda önemli bir parametre olduğu bilinmektedir. Bu parametreninde etkileri araştırılmış ve pratik su enjeksiyon uygulamalarında nasıl göz "önüne alınacağı gösterilmeye çalışılmıştır. Bunlara ek olarak, hareketlilik (möbilite) oranının yüksek olması ile viskoz parmaklaşma şeklinde ortaya çıkan dengesiz akışın heterojenlik ile nasıl etkilendiği incelenmeye çalışılmış, değişimi düşük ve yüksek, korelasyonlu ve korelasyonsuz geçirgenlik değerleri için örnekler gösterilmiştir. Geçirgenlik değişimi düşük ve korelasyonsuz sistemler ortalama değer ile sistemi temsil eden homojen ortam ile benzer sonuçlar vermesine karşın, korelasyonlu ve değişimi yüksek sistemlerde aynı durum için oldukça farklı sonuçlar elde edilmektedir. Bu çalışma ile, su enjeksiyonu performans değerlendirmesinde rezervuar karakterizasyonunun çok büyük önem taşıdığı ve rezervuar heterojenliğinin mümkün olduğunca iyi tanımlanmasının gerekliliği ve rezervuar sayısal simülasyon modelleri ile su enjeksiyon performansının tasarımın daha iyi yapılabileceği gösterilmeye çalışılmıştır. Xİİİ EFFECTS OF RESERVOIR HETEROGENEITY ON WATERFLOODING PERFORMANCE SUMMARY Primary recovery can yield % 15-40 of the orginal oil in place (OOIP), leaving rest of the OOIP unrecovered. This production can get even worse by the effects of other factors in the reservoir. Therefore, the main goal of the Enhanced Oil Recovery (EOR) methods has been to provide further increase in oil recovery. In this methods, mechanical energy by the injection of water and gas, thermal energy by the injection of hot water, steam, and in situ combustion, and chemical energy by the injection of diluted chemical fluids are provided into the reservoir. The effectiveness of Enhanced Oil Recovery methods depends on, 1. The characteristics of the reservoir (depth, structure, homogeneity, petrophysical properties). 2. Properties of displacing and displaced fluids 3. The placement of the injection and production wells. The location of the injection wells depends on the geological structure and the volume of the hydrocarbon formation. Two strategies are applied in the selection of the well locations. The first one is to displacement with patterns, and the other one is central and peripheral displacement. The estimation of the production performance is very important in waterflooding. Simplified analytical or approximate solutions are developed for prediction of recovery in systems that the injection is carried out with patterns. When there is complicated flow or boundary conditions, high viscosity ratios, and heterogeneity, the applicability of these methods will no longer be valid. Therefore, recovery performance that includes all these conditions can be done with numerical modeling. Then, it would be possible to estimate the amount of remaining oil and the regions that contain high oil saturations. This information is also valuable for infill drilling operations. Wettability and capillary pressure are very important parameters in porous medium and can not be neglected on the recovery performance predictions. Simplified analytical approaches neglect these factors. For these reasons, numerical modeling is often used for the estimation of recovery performance by waterflooding. In this study, waterflooding was preferred in order to increase the reservoir pressure and to provide an increase in oil production. Since waterflooding has been widely and economically used all over the World as an effective Enhanced Oil Recovery method. Furthermore, BOAST II, a numerical modeling method developed by the U.S Department of Energy, was chosen to study on waterflooding performance calculations. The second chapter of this study covers the various approaches that have been reported in the literature to resolve the problem of reservoir heterogeneity and its effects on waterflooding performance. In the third chapter, analytical approaches used in the evaluation of waterflooding performance are reviewed, and reservoir heterogeneity is described. Before the injection of water, the course of waterflooding must be designed. There

Primary recovery can yield % 15-40 of the orginal oil in place (OOIP), leaving rest of the OOIP unrecovered. This production can get even worse by the effects of other factors in the reservoir. Therefore, the main goal of the Enhanced Oil Recovery (EOR) methods has been to provide further increase in oil recovery. In this methods, mechanical energy by the injection of water and gas, thermal energy by the injection of hot water, steam, and in situ combustion, and chemical energy by the injection of diluted chemical fluids are provided into the reservoir. The effectiveness of Enhanced Oil Recovery methods depends on, 1. The characteristics of the reservoir (depth, structure, homogeneity, petrophysical properties). 2. Properties of displacing and displaced fluids 3. The placement of the injection and production wells. The location of the injection wells depends on the geological structure and the volume of the hydrocarbon formation. Two strategies are applied in the selection of the well locations. The first one is to displacement with patterns, and the other one is central and peripheral displacement. The estimation of the production performance is very important in waterflooding. Simplified analytical or approximate solutions are developed for prediction of recovery in systems that the injection is carried out with patterns. When there is complicated flow or boundary conditions, high viscosity ratios, and heterogeneity, the applicability of these methods will no longer be valid. Therefore, recovery performance that includes all these conditions can be done with numerical modeling. Then, it would be possible to estimate the amount of remaining oil and the regions that contain high oil saturations. This information is also valuable for infill drilling operations. Wettability and capillary pressure are very important parameters in porous medium and can not be neglected on the recovery performance predictions. Simplified analytical approaches neglect these factors. For these reasons, numerical modeling is often used for the estimation of recovery performance by waterflooding. In this study, waterflooding was preferred in order to increase the reservoir pressure and to provide an increase in oil production. Since waterflooding has been widely and economically used all over the World as an effective Enhanced Oil Recovery method. Furthermore, BOAST II, a numerical modeling method developed by the U.S Department of Energy, was chosen to study on waterflooding performance calculations. The second chapter of this study covers the various approaches that have been reported in the literature to resolve the problem of reservoir heterogeneity and its effects on waterflooding performance. In the third chapter, analytical approaches used in the evaluation of waterflooding performance are reviewed, and reservoir heterogeneity is described. Before the injection of water, the course of waterflooding must be designed. There xiv are two main factors in a design of a waterflooding. These are economical and technical factors. In a design of a waterflooding, the following steps must be considered: a ) Evaluation of the reservoir b ) Choice of a potential displacement plan c ) Estimation of the production and injection rate d) Description of the reservoir variables The purpose of the reservoir description for the design of the waterflooding is followed as: - Description of the vertical and areal size of the reservoir - Estimation of the rock properties as permeability and porosity - Estimation of the oil source distribution in the reservoir - Definition of the required fluid properties for the estimation of waterflooding performance. The most important parameter in the evaluation of waterflooding performance is the reservoir heterogeneity that describes the variation of reservoir parameters with location. It can be described as variations of areal, vertical directions, and fractures. Ideally, if the reservoir is homogeneous, a measured value from one location in the reservoir describes all the reservoir. A simulation of the homogeneous reservoir is easier than that of the heterogeneous reservoir. If the reservoir is heterogeneous, reservoir properties vary from one location to another in the reservoir. These properties are the porosity, permeability, thickness, wettability, initial saturation values, and the distribution of fluid properties. Heterogeneity distribution must be described accurately in order to evaluate the reservoir heterogeneity. In the nature, most of the reservoirs shows a log-normal permeability distribution. A log-normal distribution is described by Dykstra-Parsons coefficient that is given by; V=k50-kM.l=1_en (S]) In Equation (S. 1), V is between 0 and 1. If a equals zero, V is equal to zero. This means that the reservoir system is homogeneous. If a goes to infinity, V is equal to 1, and the reservoir is heterogeneous. In order to describe heterogeneous distributions of the reservoir parameters more accurately, correlation length and the variation values must be known. In the fourth chapter, the differential equations that describe fluid flow and the numerical solutions along with the use of computer programs are given. BOAST II program that is used to study on waterflooding performance calculations simulates isothermal Darcy flow in three dimensions. It assumes reservoir fluids can be described by three fluid phases (oil, gas, and water) of constant composition with physical properties that depend on pressure only. These reservoir fluid approximations are acceptable for a large percentage of the world' s oil and gas reservoirs. BOAST II can simulate oil and/or gas recovery by fluid expansion, displacement, gravity drainage and capillary imbibition mechanisms. BOAST II is a finite-difference and IMPES (implicit pressure/explicit saturation) numerical simulator. It contains both direct and iterative solution techniques for solving systems xv of algebraic equations. The well model in BOAST II allows specification of rate or pressure constraints on well performance, and the user İs free to add or recomplete wells during the simulation. Multiple rock PVT regions may be defined, and three aquifer models are available as options. BOAST II contains flexible initialization capabilities, a bubble point tracking scheme, an automatic time step control method, and a material balance check on solution stability. This study assumes that, there are two immissible fluids in the reservoir. Therefore, transmissibility is equal to zero between grid blocks. Gas phase is neglected, and there is no flow at the outer boundary of the reservoir. For each phase, fluid flow equations are given that, For oil For water vKX.0Vd)0 B",rawVd>w B w (S.2) (S.3) For gas VK X9V% RsAVOn FUX^VO, Br B, o ? *sw'vw T ^w B w Pg* d_ öt (p zr~ + - ~ + B, Br B w J (S.4) In Equations (S.2)-(S.4), the pressures of oil, water, and gas phases are organized in a very complex problem. For many state, the difference between phase pressures is less than its differential phase potentials. In flow equations, phase pressures and potentials can be simplified by using capillary pressure concept. and P =P -P p =p -P (S.5) (S.6) Water and gas potentials are written by using Equation (S.5) and (S.6). (S.7) and O =P -P 144 (S.8) If Equations (S.7) and (S.8) combine with Equations (S.3)-(S.5) and rearrange afterwards, in this way, pressure equation will be found. xvi (B0-R»Bfl|v.K.|a-VP0+CG0-^- +(Bw-RswBg|v.K.|^VP0+CGw- +b, VK v.Bg Bo Bw y VP0 + CGg- Pg* ' öt (S.9) Gravity and capillary terms of each phase pressures are represented as CG0 CGw,andCgg a >\ CG0 = -VK IB o.144, (S.10) CG", = -VK VBW.144 c- CS.11) CG" = V K LBg P. - V 144 J RşO^O y[ P0Z Bn ll44. r^sw^w yJ ? Pc~ + 144, (S.12) P0 is calculated implicitly in Equation (S.9). Thus, So and Sw are calculated by using Equations (S.2)-(S.4). In the fifth chapter, comparison of numerical model with Buckley-Leverett analytical solution is presented. The effects of variance of permeability, correlation length, capillarity, and viscosity ratio are discussed. Oil recovery by waterflooding in layered systems is estimated, and the effecting parameters are given. Finally conclusions and suggestions are presented. In recent years, numerical simulators are frequently used to estimate the reservoir performance. In this study, BOAST II numerical simulator is used to investigate the effects of reservoir heterogeneity, capillary forces, and viscosity ratio on waterflooding performance. First of all, results obtained by BOAST II are compared with Buckley- Leverett analytical solution for increasing the confidence of numerical model. Fitting parameters (time step, grid block length etc.) are established in comparison with using 2D and 3D calculations. xvn For 2D and 3D systems, investigations is performed at a quarter of five-spot pattern. The effects of permeability variation, correlation length, capillarity, and viscosity ratio are discussed. In one dimensional systems, the effects of capillarity and heterogeneity are investigated. Results showed that, oil production increases with decreasing flow rate. On the other hand, the effects of deterministic permeability and porosity distribution, viscosity ratio, statistical permeability distribution, and capillarity are investigated on waterflooding performance in two dimensional systems. In 2D system, the production performance of a low permeability region (1 md) between injection well and production well is better than that of a high permeability (2000 md) region. Breakthrough time was late because of the low permeability region. Therefore, at this patern, oil production performance is better. According to this, more high permeability (200 md) region where is around the low permeability region sweeps much oil. In the same system, permeability value (200 md) is fixed in all reservoir, and if porosity (0.25) is twice increased and decreased, the oil production performance of a low porosity region (0. 1 25) where is the middle of the 2D system is better than that of a high porosity (0.5) region. In 2D system, having a low variance (V = 0.2) and a log normal permeability distribution, waterflooding performance in heterogen region is equivalent displacement performance in homogen system having a geometric average permeability value. Whatever is log-normal distribution, if correlation length is low, that is, log normal distribution is completely random in heterogen system, displacement performance could not be influenced by heterogeneity. However, in high log-normal variance and correlation length systems, heterogen distribution can be affected rather more to waterflooding performance. For capillary forces in system, specially correlated heterogen and low oil viscosity systems affect the oil production performance positively. Therefore, water injection rate must be decrased in this kind of system, because this rate can be controlled more easily. In this study, another result is that water saturation profiles give information about reservoir heterogenity. Moreover, from water saturation profiles, the location of residual oil can be determined, and infill drilling can be performed at this region so that oil production performance can be increased.

Primary recovery can yield % 15-40 of the orginal oil in place (OOIP), leaving rest of the OOIP unrecovered. This production can get even worse by the effects of other factors in the reservoir. Therefore, the main goal of the Enhanced Oil Recovery (EOR) methods has been to provide further increase in oil recovery. In this methods, mechanical energy by the injection of water and gas, thermal energy by the injection of hot water, steam, and in situ combustion, and chemical energy by the injection of diluted chemical fluids are provided into the reservoir. The effectiveness of Enhanced Oil Recovery methods depends on, 1. The characteristics of the reservoir (depth, structure, homogeneity, petrophysical properties). 2. Properties of displacing and displaced fluids 3. The placement of the injection and production wells. The location of the injection wells depends on the geological structure and the volume of the hydrocarbon formation. Two strategies are applied in the selection of the well locations. The first one is to displacement with patterns, and the other one is central and peripheral displacement. The estimation of the production performance is very important in waterflooding. Simplified analytical or approximate solutions are developed for prediction of recovery in systems that the injection is carried out with patterns. When there is complicated flow or boundary conditions, high viscosity ratios, and heterogeneity, the applicability of these methods will no longer be valid. Therefore, recovery performance that includes all these conditions can be done with numerical modeling. Then, it would be possible to estimate the amount of remaining oil and the regions that contain high oil saturations. This information is also valuable for infill drilling operations. Wettability and capillary pressure are very important parameters in porous medium and can not be neglected on the recovery performance predictions. Simplified analytical approaches neglect these factors. For these reasons, numerical modeling is often used for the estimation of recovery performance by waterflooding. In this study, waterflooding was preferred in order to increase the reservoir pressure and to provide an increase in oil production. Since waterflooding has been widely and economically used all over the World as an effective Enhanced Oil Recovery method. Furthermore, BOAST II, a numerical modeling method developed by the U.S Department of Energy, was chosen to study on waterflooding performance calculations. The second chapter of this study covers the various approaches that have been reported in the literature to resolve the problem of reservoir heterogeneity and its effects on waterflooding performance. In the third chapter, analytical approaches used in the evaluation of waterflooding performance are reviewed, and reservoir heterogeneity is described. Before the injection of water, the course of waterflooding must be designed. There xiv are two main factors in a design of a waterflooding. These are economical and technical factors. In a design of a waterflooding, the following steps must be considered: a ) Evaluation of the reservoir b ) Choice of a potential displacement plan c ) Estimation of the production and injection rate d) Description of the reservoir variables The purpose of the reservoir description for the design of the waterflooding is followed as: - Description of the vertical and areal size of the reservoir - Estimation of the rock properties as permeability and porosity - Estimation of the oil source distribution in the reservoir - Definition of the required fluid properties for the estimation of waterflooding performance. The most important parameter in the evaluation of waterflooding performance is the reservoir heterogeneity that describes the variation of reservoir parameters with location. It can be described as variations of areal, vertical directions, and fractures. Ideally, if the reservoir is homogeneous, a measured value from one location in the reservoir describes all the reservoir. A simulation of the homogeneous reservoir is easier than that of the heterogeneous reservoir. If the reservoir is heterogeneous, reservoir properties vary from one location to another in the reservoir. These properties are the porosity, permeability, thickness, wettability, initial saturation values, and the distribution of fluid properties. Heterogeneity distribution must be described accurately in order to evaluate the reservoir heterogeneity. In the nature, most of the reservoirs shows a log-normal permeability distribution. A log-normal distribution is described by Dykstra-Parsons coefficient that is given by; V=k50-kM.l=1_en (S]) In Equation (S. 1), V is between 0 and 1. If a equals zero, V is equal to zero. This means that the reservoir system is homogeneous. If a goes to infinity, V is equal to 1, and the reservoir is heterogeneous. In order to describe heterogeneous distributions of the reservoir parameters more accurately, correlation length and the variation values must be known. In the fourth chapter, the differential equations that describe fluid flow and the numerical solutions along with the use of computer programs are given. BOAST II program that is used to study on waterflooding performance calculations simulates isothermal Darcy flow in three dimensions. It assumes reservoir fluids can be described by three fluid phases (oil, gas, and water) of constant composition with physical properties that depend on pressure only. These reservoir fluid approximations are acceptable for a large percentage of the world' s oil and gas reservoirs. BOAST II can simulate oil and/or gas recovery by fluid expansion, displacement, gravity drainage and capillary imbibition mechanisms. BOAST II is a finite-difference and IMPES (implicit pressure/explicit saturation) numerical simulator. It contains both direct and iterative solution techniques for solving systems xv of algebraic equations. The well model in BOAST II allows specification of rate or pressure constraints on well performance, and the user İs free to add or recomplete wells during the simulation. Multiple rock PVT regions may be defined, and three aquifer models are available as options. BOAST II contains flexible initialization capabilities, a bubble point tracking scheme, an automatic time step control method, and a material balance check on solution stability. This study assumes that, there are two immissible fluids in the reservoir. Therefore, transmissibility is equal to zero between grid blocks. Gas phase is neglected, and there is no flow at the outer boundary of the reservoir. For each phase, fluid flow equations are given that, For oil For water vKX.0Vd)0 B",rawVd>w B w (S.2) (S.3) For gas VK X9V% RsAVOn FUX^VO, Br B, o ? *sw'vw T ^w B w Pg* d_ öt (p zr~ + - ~ + B, Br B w J (S.4) In Equations (S.2)-(S.4), the pressures of oil, water, and gas phases are organized in a very complex problem. For many state, the difference between phase pressures is less than its differential phase potentials. In flow equations, phase pressures and potentials can be simplified by using capillary pressure concept. and P =P -P p =p -P (S.5) (S.6) Water and gas potentials are written by using Equation (S.5) and (S.6). (S.7) and O =P -P 144 (S.8) If Equations (S.7) and (S.8) combine with Equations (S.3)-(S.5) and rearrange afterwards, in this way, pressure equation will be found. xvi (B0-R»Bfl|v.K.|a-VP0+CG0-^- +(Bw-RswBg|v.K.|^VP0+CGw- +b, VK v.Bg Bo Bw y VP0 + CGg- Pg* ' öt (S.9) Gravity and capillary terms of each phase pressures are represented as CG0 CGw,andCgg a >\ CG0 = -VK IB o.144, (S.10) CG", = -VK VBW.144 c- CS.11) CG" = V K LBg P. - V 144 J RşO^O y[ P0Z Bn ll44. r^sw^w yJ ? Pc~ + 144, (S.12) P0 is calculated implicitly in Equation (S.9). Thus, So and Sw are calculated by using Equations (S.2)-(S.4). In the fifth chapter, comparison of numerical model with Buckley-Leverett analytical solution is presented. The effects of variance of permeability, correlation length, capillarity, and viscosity ratio are discussed. Oil recovery by waterflooding in layered systems is estimated, and the effecting parameters are given. Finally conclusions and suggestions are presented. In recent years, numerical simulators are frequently used to estimate the reservoir performance. In this study, BOAST II numerical simulator is used to investigate the effects of reservoir heterogeneity, capillary forces, and viscosity ratio on waterflooding performance. First of all, results obtained by BOAST II are compared with Buckley- Leverett analytical solution for increasing the confidence of numerical model. Fitting parameters (time step, grid block length etc.) are established in comparison with using 2D and 3D calculations. xvn For 2D and 3D systems, investigations is performed at a quarter of five-spot pattern. The effects of permeability variation, correlation length, capillarity, and viscosity ratio are discussed. In one dimensional systems, the effects of capillarity and heterogeneity are investigated. Results showed that, oil production increases with decreasing flow rate. On the other hand, the effects of deterministic permeability and porosity distribution, viscosity ratio, statistical permeability distribution, and capillarity are investigated on waterflooding performance in two dimensional systems. In 2D system, the production performance of a low permeability region (1 md) between injection well and production well is better than that of a high permeability (2000 md) region. Breakthrough time was late because of the low permeability region. Therefore, at this patern, oil production performance is better. According to this, more high permeability (200 md) region where is around the low permeability region sweeps much oil. In the same system, permeability value (200 md) is fixed in all reservoir, and if porosity (0.25) is twice increased and decreased, the oil production performance of a low porosity region (0. 1 25) where is the middle of the 2D system is better than that of a high porosity (0.5) region. In 2D system, having a low variance (V = 0.2) and a log normal permeability distribution, waterflooding performance in heterogen region is equivalent displacement performance in homogen system having a geometric average permeability value. Whatever is log-normal distribution, if correlation length is low, that is, log normal distribution is completely random in heterogen system, displacement performance could not be influenced by heterogeneity. However, in high log-normal variance and correlation length systems, heterogen distribution can be affected rather more to waterflooding performance. For capillary forces in system, specially correlated heterogen and low oil viscosity systems affect the oil production performance positively. Therefore, water injection rate must be decrased in this kind of system, because this rate can be controlled more easily. In this study, another result is that water saturation profiles give information about reservoir heterogenity. Moreover, from water saturation profiles, the location of residual oil can be determined, and infill drilling can be performed at this region so that oil production performance can be increased.

##### Açıklama

Tez (Yüksek Lisans) -- İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, 1996

Thesis (M.Sc.) -- İstanbul Technical University, Institute of Science and Technology, 1996

Thesis (M.Sc.) -- İstanbul Technical University, Institute of Science and Technology, 1996

##### Anahtar kelimeler

Rezervasyon,
Su enjeksiyonu,
Reservation,
Water injection