Reservoir History Matching And Injection-production Program Adjustment Of Block Mao 733 In Aonan Oilfield

The Aonan Oilfield is located in the central depression of the northern Songliao Basin and in the joint of three secondary structural zones-Daqing Changyuan,Gulong Depression,Shanzhao Depression.Mao 733 Block of Aonan Oilfield is located in a pure oil zone,which is a small fault block stretched in south-north tectonically.Reservoir sand are external delta front deposits with mainly sheet sand and backshore sand ridge,sand body of main oil reservoirs of PⅠ1-PⅠ4 is developed stably,and sand body distribution is overall thick in the north and thin in the south,and the oil potential is larger,but the sand body is very scattered and the thickness is thin overally,the porosity and permeability property is not that good.After years of underground exploitation in Mao 733 Block of Aonan Oilfield,the original well pattern exploitation effect gradually deteriorated.It is difficult to establish an effective displacment under the condition of the current well pattern in the low permeability oilfield.Therefore,it is necessary to study the reasonable comprehensive injection and production adjustment technology and method to provide the basis for the later scale application,while establishing a reasonable reservoir description model is its premise,in this paper,a reservoir simulation automatic history matching case is proposed to more objectively and more efficiently obtain the reservoir model which conforms to the underground fluid flow.According to the drilling geological information,well stratified data,single well deviation data,interlayer data,breakpoint data,small layer data and so on,for reservoir of PⅠ1-PⅠ4 in aonan oilfield Mao 733 block structure model,sand body model,and property model has been successively established,three dimensional geological model of 109x184x4 grid cells can be obtained,through analysis of geological model,we can know that the maximum porosity in Putaohua reservoir in the block is 17.67%,a minimum of 5.5%,and an average of 12.99%;the maximum permeability is 1.68x10-3μm2,a minimum of 0.067x10-3μm2,and an average of 0.53×10-3μm2.which indicates a midddle-low porosity,low,super-low permeability reservoir;the maximum water saturation is up to 64%,a minimum of 25%,and an average of 54%,indicating a higher critical water satuation int the block.Reserves matching should be done before histroy matching,through uncertainty and sensitivity analysis of reserves,influence of NTG on reserves is biggest,that of porosity is secondary,that of water saturation is smallest.We choose PROV multiplier,a combination parameter of porosity and NTG to adjust the reserves,based on the individual layer reserve distribution curve,we optimize individual reservoir layer of PⅠ1-PⅠ4 the PROV multiplier of 1.06,1.17,2.15,0.69 respectively,and thus we get individual layer reserve of ***× 10~4m~3,***× 10~4m~3,***× 10~4m~3,***× 10~4m~3,the total relative error against the block proven reserves is-0.026%.In the softwarer MEPO,the upper and lower limit,distribution function of uncertainty parameters,such as infiltration curve parameters are set as input parameters,output parameters such simulation liquid production and oberserved liquid production are set as response parameters to construct objective function and then sensitivity algorithm,latin hypercube algorithm and genetic algorithem are performed successively to obtain the final best matching model to perform automatic history matching.Sensitivity algorithm is used by transforming the upper and lower limits of different uncertainty parameters to generate 57 models,analysis of the effects of uncertainty parameters on the global value,field liquid production rate,field oil production rate,water cut and single well oil production rate,water cut indicates that the effects of infiltration curve parameters on the global value are biggest,effects of saturation parameters are secondary,and effects of the permeability parameters are smallest;based on the uncertainty parameter sensitivity analysis,the upper and lower limits of individual uncertainty parameters are adjusted.Latin hypercube algorithm scans algorithm search space through stratified random sampling of uncertainty pareameters,200 models were generated,and we select 72 experiments with global minimum as initial population of genetic algorithm optimization in the next step,and Latin hypercube algorithm analysis indicates that vertical permeability multiplier has very little relationships with global value and production rate,water cut,and thus are removed from the uncertainty parameter sets.Genetic algorithm optimization terminated after 22 times iteration,that is,the relative error between the global values of the last 10 generations is less than 5%,and a total of 504 models were generated.The minimal global value of these models is 5820.634,the target model is the model with the global value of the minimal global value.As we has changed the initial water saturation,and the reserve need to be adjusted again,we optimize individual reservoir layer of PⅠ1-PⅠ4 the PROV multiplier of 0.98,1.12,2.5,0.57 respectively,and thus we get individual layer reserve of ***× 10~4m~3,***× 10~4m~3,***× 10~4m~3,***× 10~4m~3,the total relative error against the block proven reserves is-0.029%.The matching of the field liquid production rate of the best optimization model is very good,the average absolute error between block simulation production rate and history production rate is 0.10 m3/d,the maximum absolute error is-1.63m3/d,but the absolute error in more than two-third of the time step is 0.0001 m3/d;Block oil production rates are well matched for most of the time step,and the average absolute error between block simulation oil production rate and history block oil production rate is 0.14 m3/d,the maximum absolute error of 1.26 m3/d,Because the water cut in this block is abnormal with discontinuous pulse water cut,and the block water cut matching is slightly worse.However,the shape of the curve peak of water cut in most of the time scale is matched.The absolute error between block simulation water cut and history water cut is-0.012,the maximum absolute error is-0.12;Single well oil production rate matching is very good,the water cut is poor matching,the shape of the curve peak is more or less matched.As to injection and production program adjustment,based on the injection intensity and injection cycles,21 injection programs are designed and different programs of timimg in water cut when production wells turn to injection well.Based on the analyis of different injection program,we can know that in the case of the same injection intensity of the injection and oil-to-injection wells,the smaller injection intensity is better;in the case that injection intensity of the injection wells is controlled,injection intensity of oil-to-injection wells is strengthened,smaller injection intensity for oil-to-injection wells is still better.Periodic water injection has more advantages relative to the non-periodic water injection at some extent,alternating cycle of water injection has the advantages against the synchronous cycle water injection;in the case of the same water injection intensity,the alternating cycle of water injection that the oil-to-injection wells open in one cycle and close in two cycles,the injection well close in two cycles and open in one cycle is the best.For Aonan Mao 733 block,the 20 th water injection program is the most optimal,the timimg when the water cut is 10% is the best timing for production well to injection well.