An evaluation method for tight oil

Abstract

This invention relates to a method for evaluating tight oil, belonging to the field of oil and gas exploration technology. This evaluation method is used when pilot development data for the target area is incomplete or no pilot development test has been conducted. First, it characterizes the "sweetness" of the tight oil by using the ratio of the stable daily production per kilometer depth in the early stages of single-well production to the lower limit of the daily production of industrial oil and gas flow per kilometer depth. Then, it uses the sweetness of the tight oil "sweetness" from the same basin as the target area and the attribute values ​​of various geophysical parameters as the output and input of a neural network, respectively, to train the model and obtain a neural network relationship model representing the relationship between sweetness and the attribute values ​​of the geophysical parameters. Using this neural network relationship model, combined with the normalized attribute values ​​of the geophysical parameters of the tight oil "sweetness" to be evaluated, the sweetness value of the tight oil "sweetness" to be evaluated is predicted with high accuracy, enabling a quantitative evaluation of tight oil production capacity.

Description

Technical Field

[0001] This invention belongs to the field of oil and gas exploration technology, and specifically relates to a method for evaluating tight oil. Background Technology

[0002] Tight oil refers to oil accumulations formed within or adjacent to high-quality source rock formations, such as tight clastic rocks, shale, or carbonate rocks, without long-distance migration. Generally, it has no natural production capacity, or its natural production capacity is below the economic flow threshold, requiring large-scale modification to achieve industrial production. The "sweet spot" of tight oil refers to a range of tight oil reservoirs with a favorable configuration of source rock, reservoir, and engineering (reservoir mechanics) quality, where potential development value can be obtained through reservoir modification. In existing technologies, the concept of "sweetness" has been proposed to characterize the quality of tight oil "sweet spots." Sweetness is generally considered a parameter characterizing the degree of sweetness, a quantitative evaluation index of sweet spot production capacity, and is positively correlated with post-pressurization production capacity.

[0003] In their paper titled "Research on the Evaluation Method of Sweet Spot in Tight Oil," published in the second issue of the journal *Special Oil and Gas Reservoirs* in 2017, authors Chen Fuli et al. proposed a method for evaluating the sweet spot of tight oil. This method uses the lower limit of the internal rate of return (IRR) determined by tight oil development enterprises as a reference standard, calculating the "sweet spot" sweetness by the ratio of the actual revenue from tight oil development to the lower limit of the IRR standard. A calculation model for the "sweet spot" sweetness of tight oil is established, and a sweet spot evaluation chart is created. Based on the evaluation results of the chart, tight oil areas with reasonable sweetness are selected for production capacity construction in a timely manner. This method requires a large amount of pilot test data from development wells in the same tight oil sweet spot reservoir. Based on the production characteristics of single wells in the tight oil development test area or development area, evaluation methods such as decline analysis can be used to determine the cumulative oil production when a single well is abandoned, thereby evaluating the sweetness of the tight oil.

[0004] The main problem with this method is its narrow scope of application and strong limitations, which are due to the following reasons:

[0005] First, the sweetness evaluation must be performed on blocks that have received investment and returns. Blocks that have not received investment and returns cannot have their "sweetness" predicted.

[0006] Secondly, the scale of a single tight oil resource requires pilot testing with more than 30 development wells, which is demanding and involves a large workload. When there are fewer than 30 development wells available for pilot testing, the representativeness of the sweetness evaluation is poor, leading to inaccurate sweetness prediction. This results in the selection of an unsuitable tight oil reservoir for production capacity construction, which in turn affects economic benefits.

[0007] The paper titled "A New Technology for Shale Gas Reservoir Evaluation—Sweetness Evaluation Method," published in the 4th issue of the journal *Petroleum Drilling Technology* in 2016, by authors Jiang Tingxue et al., proposes the concepts of geological sweetness and engineering sweetness for shale gas, and precisely characterizes and quantifies the "sweetness" of these two parameters. A benchmark is set for the sweetest geological and engineering sweetness in the same evaluation area, encompassing a set of optimal geological and engineering parameters. Then, Euclidean proximity is calculated to characterize the similarity between the parameter set of the evaluation area and this benchmark, and this similarity is used as the geological and engineering sweetness. After determining the weight allocation of geological and engineering sweetness using a grey relational analysis method, a comprehensive sweetness index is calculated to optimize the location of fracturing clusters in horizontal shale gas wells. This method mainly has the following two problems:

[0008] First, it is applicable to different sections of the same well or different wells in the same block, which limits its applicability.

[0009] Secondly, the highest sweetness benchmark needs to be selected. In practice, there is a possibility that the selected benchmark is lower than the unknown range or well, which may cause a dessert with high sweetness to be misjudged as low sweetness, affecting the dessert quality assessment.

[0010] Furthermore, since domestic exploration of tight oil "sweet spots" mostly employs vertical well fracturing, while shale gas exploration primarily utilizes horizontal well volumetric fracturing, and these methods are applicable to shale gas reservoirs with the same reservoir structure, accurately predicting newly discovered small tight oil "sweet spots" without pilot testing remains a pressing issue. Summary of the Invention

[0011] The purpose of this invention is to provide a method for evaluating dense oils, which addresses the problem of inaccurate evaluation of the "sweetness" of dense oils in existing methods.

[0012] Based on the above objectives, the technical solution for an evaluation method of tight oil is as follows:

[0013] (1) Obtain several tight oil "sweet spots" samples that have undergone pilot development tests and belong to the same basin as the tight oil reservoir to be evaluated, and determine the sweetness value and geophysical parameter attribute values ​​of each tight oil "sweet spot"; wherein, the sweetness value of the tight oil "sweet spot" is determined based on the ratio of the stable daily production per kilometer depth in the early stage of single well production to the lower limit of the daily production of industrial oil and gas flow per kilometer depth, and the geophysical parameter attribute values ​​include: instantaneous amplitude, continuity, instantaneous frequency and instantaneous phase;

[0014] (2) Normalize the attribute values ​​of each geophysical parameter in each dense oil “sweet spot”;

[0015] (3) The attribute values ​​of each geophysical parameter of each tight oil "sweet spot" after normalization are used as the input of the neural network, and the sweetness value of each tight oil "sweet spot" is used as the output of the neural network. The model is trained to determine the neural network relationship model that represents the relationship between the sweetness value of the tight oil "sweet spot" and the attribute values ​​of each geophysical parameter.

[0016] (4) Obtain the attribute values ​​of each attribute in the geophysical parameters of the tight oil “sweet spot” to be evaluated after normalization, and substitute them into the neural network relationship model to determine the sweetness of the tight oil “sweet spot” to be evaluated.

[0017] The beneficial effects of the above technical solution are:

[0018] The evaluation method of this invention is applicable when pilot development data for the target area is incomplete or no pilot development test has been conducted. First, the ratio of the stable daily production per kilometer depth in the early stages of single-well production to the lower limit of the daily production of industrial oil and gas at a kilometer depth is used to reasonably characterize the sweetness of tight oil (i.e., the sweetness of the "sweet spot"). Then, using the sweetness of the tight oil "sweet spot" from the same basin as the target area and the attribute values ​​of various geophysical parameters, which have already undergone pilot development tests, as the output and input of a neural network, a neural network relationship model is obtained to represent the relationship between the sweetness of the tight oil "sweet spot" and the attribute values ​​of the geophysical parameters. Using this neural network relationship model, combined with the normalized attribute values ​​of the geophysical parameters of the tight oil "sweet spot" to be evaluated, the sweetness value of the tight oil "sweet spot" to be evaluated is predicted with high accuracy. This allows for a quantitative evaluation of tight oil production capacity, thereby achieving the goal of optimizing the exploration target for the benefits of the "sweet spot".

[0019] Furthermore, in order to determine the sweetness value of the reservoir, the sweetness value of the tight oil "sweetness" is calculated using the following formula:

[0020]

[0021] In the formula, T is the sweetness value of the dense oil "dessert", and x i The value represents the stable daily oil production per kilometer depth of the i-th well within the tight oil "sweet spot" during the initial production phase, where i = 1, 2, ..., n, and n is the number of individual wells within the tight oil "sweet spot"; k is the lower limit of the daily production of industrial oil and gas flow per kilometer depth.

[0022] Furthermore, to ensure the representativeness of the obtained geophysical parameter attribute values, when the area of ​​the tight oil "sweet spot" is greater than a set value, the average value of the geophysical parameters at all well points within the tight oil "sweet spot" is taken as the final attribute value of the tight oil "sweet spot"; when the area of ​​the tight oil "sweet spot" is not greater than the set value, the attribute value of the geophysical parameters at a single well point within the tight oil "sweet spot" is taken as the final attribute value of the tight oil "sweet spot".

[0023] Furthermore, to facilitate rapid training of the neural network and to standardize the units of measurement, it is necessary to normalize the attribute values ​​of the geophysical parameters. The calculation formula for normalizing the attribute values ​​is as follows:

[0024] vv = (v - vmin) / (vmax - vmin)

[0025] In the formula, v is the attribute value of the geophysical parameter before normalization, vv is the attribute value of the geophysical parameter after normalization, and vmin and vmax are the minimum and maximum attribute values ​​of the seismic data of the same stratum of the basin to which the tight oil "sweet spot" to be evaluated belongs, respectively. Attached Figure Description

[0026] Figure 1 This is a flowchart of the evaluation method for tight oil in an embodiment of the present invention. Detailed Implementation

[0027] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0028] This embodiment proposes a method for evaluating the sweetness of dense oils, specifically for assessing the sweetness of dense oil "desserts". The basic idea is as follows:

[0029] The ratio of the relatively stable daily production per kilometer depth in the early stage of a single well to the lower limit of the daily production of industrial oil and gas at a kilometer depth is used as an indicator to judge the sweetness. Geophysical parameters of "sweet spots" that have already been developed for pilot tests within the same basin are extracted and added to the input end of a neural network. The sweetness of the "sweet spots" that have already been developed for pilot tests is input to the output end of the neural network. Then, the model is trained according to the neural network algorithm to establish a neural network relationship model between the input parameters and the output parameters. This relationship model is then applied to the geophysical parameters of several "sweet spots" to be evaluated to obtain the sweetness value of the "sweet spots". By ranking the sweetness of the "sweet spots", the optimal exploration targets for the benefits of "sweet spots" can be selected.

[0030] The overall process of the above method is as follows: Figure 1 As shown, the specific implementation steps are as follows:

[0031] Step 1: Collect multiple "sweet spots" samples that have undergone pilot development tests and belong to the same basin as the "sweet spots" to be evaluated. To ensure the fitting accuracy of the relationship network between the sweetness of the "sweet spots" and geophysical parameters in the subsequent steps, at least 5 "sweet spot" samples that have undergone pilot development tests within the same basin should be selected. Therefore, in this step, 5 "sweet spots" samples that have undergone pilot development tests in a certain area of ​​a certain oilfield are selected.

[0032] Step 2, calculate the sweetness of each "dessert" sample; the formula for calculating sweetness is:

[0033]

[0034] Where T is the sweetness value, x i The first part represents the relatively stable daily oil production per kilometer depth of the i-th well within the "sweet spot" zone, where i = 1, 2, ..., n; k is the lower limit of the daily production of industrial oil and gas flow per kilometer depth within the oil-producing zone as specified in the oil and gas reserve assessment standard.

[0035] In this step, the oil-producing sections of the selected "sweet spot" samples are all in the 1000-2000 meter well range. The lower limit of the daily production of industrial oil and gas flow at a well depth of 1 kilometer is 1 cubic meter. Taking the "sweet spot" block A as an example, the "sweet spot" block A has 12 wells. The relatively stable daily production per kilometer of well depth for a single well in the early stage were 10.76 cubic meters, 10.8 cubic meters, 6.5 cubic meters, 11.78 cubic meters, 8.76 cubic meters, 7.58 cubic meters, 10.12 cubic meters, 6.85 cubic meters, 5.78 cubic meters, 6.23 cubic meters, 8.88 cubic meters, and 7.12 cubic meters. Using the sweetness calculation formula above, the sweetness value of the "sweet spot" block A is calculated to be 8.43. Then, using the sweetness calculation formula above, the sweetness values ​​of B, C, D, and E are calculated to be 1.52, 0.85, 5.70, and 10.82, respectively.

[0036] Step 3: Extract and normalize the geophysical parameters of the "sweet spot" sample. These geophysical parameters include four attribute values: instantaneous amplitude, energy (i.e., continuity), instantaneous frequency, and instantaneous phase of the seismic data. The formula for normalizing each attribute value is as follows:

[0037] vv = (v - vmin) / (vmax - vmin)

[0038] Where v is the attribute value of the geophysical parameter before normalization, vv is the attribute value of the geophysical parameter after normalization, and vmin and vmax are the minimum and maximum attribute values ​​of the seismic data of the same stratum of the basin to which the "sweet spot" to be evaluated belongs, respectively.

[0039] In this step, the four attribute values ​​v of the five "desserts" are determined, along with the minimum value vmin and maximum value vmax of the four attributes of the basin strata to which the dessert to be evaluated belongs. Then, the calculation is performed according to the normalization formula above, and the calculation results are as follows:

[0040] For the "sweet spot" A block: the normalized instantaneous amplitude is 0.82, the instantaneous frequency is 0.24, the instantaneous phase is 0.46, and the continuity is 0.67;

[0041] For the "sweet" block B: the normalized instantaneous amplitude is 0.23, the instantaneous frequency is 0.85, the instantaneous phase is 0.37, and the continuity is 0;

[0042] For the "sweet spot" C block: the normalized instantaneous amplitude is 0.02, the instantaneous frequency is 0.9, the instantaneous phase is 0.89, and the continuity is 0.12;

[0043] For the "sweet spot" D block: the normalized instantaneous amplitude is 0.47, the instantaneous frequency is 0.59, the instantaneous phase is 0.43, and the continuity is 0.48;

[0044] For the “sweet spot” E block: the normalized instantaneous amplitude is 0.95, the instantaneous frequency is 0.1, the instantaneous phase is 0.3, and the continuity is 0.9.

[0045] In this step, for areas larger than 0.5 km² 2 The "sweet spot" is determined by taking the average of the geophysical parameters at all well points within the block as the final attribute value of the sweet spot; for areas less than or equal to 0.5 km²... 2 The "dessert" is determined by taking the attribute values ​​of geophysical parameters at the well point as the final attribute value of the dessert.

[0046] Step 4: Input the attribute values ​​of the normalized geophysical parameters from Step 3 into the input terminal of the neural network, and input the sweetness of each "dessert" from Step 2 into the output terminal of the neural network; train the model according to the neural network algorithm, establish the nonlinear mapping relationship between the sweetness of the "dessert" and the geophysical parameters, and determine the neural network relationship model.

[0047] In this step, the neural network relational model has various structures. This invention uses a BP (back propagation, multi-layer feedforward) structure to establish a nonlinear mapping relationship. A BP neural network relational model consists of an input layer, one or more intermediate layers, and an output layer. Each layer contains multiple nodes, and the nodes between layers are interconnected to form a network. In step 3, the normalized geophysical parameter attribute values ​​are input as known information from the input end of the network relational model. After network computation, they reach the output layer to obtain the sweetness Y = (y1, y2, ..., yn) of the "dessert".

[0048] Therefore, the attribute values ​​of the geophysical parameters corresponding to the "sweet spot" oil-producing layer determined in step 3 are input from the input end. Five nodes are set at the output end, and the sweetness of the five "sweet spots" is controlled respectively, so that the sweetness y1 of the "sweet spot" block A is 8.43, the sweetness y2 of the block B is 1.52, the sweetness y3 of the block C is 0.85, the sweetness y4 of the block D is 5.70, and the sweetness y5 of the block E is 10.82. After training, the nonlinear mapping relationship between the network nodes is obtained, thereby determining the neural network relationship model representing the nonlinear mapping relationship.

[0049] Step 5: After determining the neural network relationship model, the sweetness of the "desserts" to be evaluated can be predicted by extracting and normalizing their geophysical parameters. For example, extracting and normalizing the geophysical parameters of four "desserts" (blocks F, G, H, and I) yields the following data:

[0050] For the "sweet spot" F block: the normalized instantaneous amplitude is 0.75, the instantaneous frequency is 0.23, the instantaneous phase is 0.13, and the continuity is 0.87;

[0051] For the "sweet spot" G block: the normalized instantaneous amplitude is 0.43, the instantaneous frequency is 0.53, the instantaneous phase is 0.65, and the continuity is 0.05;

[0052] For the "sweet spot" H block: the normalized instantaneous amplitude is 0.15, the instantaneous frequency is 0.93, the instantaneous phase is 0.78, and the continuity is 0.12;

[0053] For the "sweet" I block: the normalized instantaneous amplitude is 0.63, the instantaneous frequency is 0.32, the instantaneous phase is 0.05, and the continuity is 0.88.

[0054] Step 6: Input the attribute values ​​of the geophysical parameters obtained in Step 5 into the neural network relationship model established in Step 4 to obtain the sweetness of the "sweet spots" to be evaluated. Therefore, the sweetness values ​​of "sweet spot" blocks F, G, H, and I can be determined to be 8.21, 2.37, 1.05, and 9.74, respectively. By ranking the sweetness of the "sweet spots" to be evaluated and comparing them, it can be seen that "sweet spot" block I has the highest sweetness and is the preferred target for profitable exploration.

[0055] The evaluation method of the present invention has the following characteristics:

[0056] 1) Applicable to situations where the development pilot test data for the tight oil “sweet spot” (i.e. the “sweet spot” to be evaluated) is incomplete or no development pilot test has been conducted. Several tight oil “sweet spot” samples from the same basin that have undergone development pilot tests can be used to train a neural network to predict the sweet spot of the tight oil to be evaluated.

[0057] 2) The method of the present invention can accurately predict the sweetness value of the tight oil "sweetness" to be evaluated, and thus can reliably select the tight oil "sweetness" with the highest sweetness value for priority exploration, avoiding the misjudgment of the actual high sweetness "sweetness" as low sweetness.

[0058] 3) The ratio of the daily oil and gas production per kilometer of well depth to the lower limit of the daily industrial oil and gas production per kilometer of well depth is selected as the sweetness of the "sweet spot" evaluation standard. For oil and gas reservoirs with comparable production, when the reservoir is shallow, the drilling difficulty is lower. Using this sweetness evaluation method is beneficial to select the "sweet spots" with low drilling difficulty and improve the exploration and development effect of tight oil.

[0059] 4) By using a neural network, a neural network relationship model trained on "dessert" samples can reliably predict the sweetness of "desserts" without requiring specific physical relationships.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method of evaluating tight oil, characterized by, The method comprises the following steps: (1) obtaining several tight oil "dessert" samples which belong to the same basin as the tight oil reservoir to be evaluated and have been developed through pilot test, determining the sweetness value of each tight oil "dessert" and the attribute value of the geophysical parameter; wherein the sweetness value of the tight oil "dessert" is determined according to the ratio of the stable kilometer well depth daily production in the early stage of single well production to the lower limit value of kilometer well depth industrial oil and gas flow daily production, and the attribute value of the geophysical parameter includes instantaneous amplitude, continuity, instantaneous frequency and instantaneous phase; (2) normalizing each attribute value in the geophysical parameter of each tight oil "dessert"; (3) taking each attribute value in the geophysical parameter of each tight oil "dessert" after normalization as the input of the neural network, taking the sweetness value of each tight oil "dessert" as the output of the neural network, performing model training, and determining the neural network relationship model representing the relationship between the sweetness value of the tight oil "dessert" and each attribute value in the geophysical parameter; (4) obtaining each attribute value in the geophysical parameter of the normalized tight oil "dessert" to be evaluated, and substituting it into the neural network relationship model to determine the sweetness of the tight oil "dessert" to be evaluated.

2. The method of evaluating tight oil of claim 1, wherein, The sweetness value of the tight oil "dessert" is calculated according to the following formula: where T is the sweetness value of the tight oil "dessert", x i represents the initial stable daily oil production of the i-th well in the tight oil "dessert" per kilometer of well depth, i = 1, 2, …, n, n is the number of single wells in the tight oil "dessert"; k is the lower limit value of the daily production of industrial oil and gas stream per kilometer of well depth.

3. The method of evaluating tight oil of claim 1, wherein, When the area of the tight oil "dessert" is greater than the set value, the average value of the attribute of the geophysical parameter at all well points in the tight oil "dessert" is taken as the final attribute value of the reservoir; when the area of the tight oil "dessert" is not greater than the set value, the attribute value of the geophysical parameter at a single well point in the reservoir is taken as the final attribute value of the tight oil "dessert".

4. The method of evaluating tight oil of claim 1, wherein, The calculation formula for normalizing each attribute value is as follows: vv=(v-vmin) / (vmax-vmin) In the formula, v is the attribute value of the geophysical parameter before normalization, vv is the attribute value of the geophysical parameter after normalization, v min and v max are the minimum attribute value and the maximum attribute value of the seismic data of the same layer system in the basin to which the tight oil "dessert" to be evaluated belongs, respectively.

Images