Slurry formula performance prediction and grade optimization method for horizontal directional drilling

Abstract

The invention discloses a slurry formula performance prediction and grade optimization method for horizontal directional drilling. The method comprises the following steps: S1, slurry experiment and data acquisition based on an HDD process; s2, establishing an HDD slurry viscosity nonlinear regression model; s3, defining a slurry performance grade for HDD construction; s4, performing Monte Carlo simulation and formula space global performance scanning; s5, visualizing the result of the step S4; S6, performing HDD field formula optimization and decision making; the method provides a standardized process and decision basis for slurry preparation. The initiative'performance grade distribution cloud picture 'enables an engineer to see all possible performance results of the formula clearly at one eye, and blind trial and error are thoroughly avoided. The lowest-cost formula meeting the requirement can be quickly locked, and the material cost is saved; meanwhile, field debugging time is shortened, and the project progress is accelerated.

Description

Technical Field

[0001] This invention relates to the field of trenchless construction technology, and in particular to a method for designing and optimizing mud systems in horizontal directional drilling (HDD). Specifically, it is a scientific method that combines mathematical statistics and computer simulation to predict mud performance, classify performance levels, and visualize the optimal formulation range across the entire process. Background Technology

[0002] In horizontal directional drilling (HDD) operations, drilling mud is considered the "lifeblood" of the project, and its performance is crucial. The mud needs to be dynamically adjusted according to different geological conditions (such as clay, sand, and silt layers) and construction stages (pilot hole, pre-reaming, pullback tubing) to achieve the following core functions: stabilizing the borehole wall, carrying and suspending drill cuttings, lubricating and cooling the drill string, and reducing pullback resistance.

[0003] Currently, adjusting the on-site mud formulation for HDD (High-Density Digging) primarily relies on engineers' experience and a trial-and-error approach. Common practices include:

[0004] (1) Preliminary formulation based on experience: Based on geological reports and past experience, the amount of additives such as bentonite, polymers (such as CMC), and soda ash is initially determined.

[0005] (2) On-site testing and adjustment: Prepare small samples and measure their Marvimeter funnel viscosity and other indicators. If the requirements are not met, add or remove certain additives based on experience and re-prepare and test.

[0006] This method has the following drawbacks:

[0007] (1) High dependence on experience: experienced engineers are scarce, and different engineers have different judgment criteria, resulting in poor formula stability.

[0008] (2) Blindness and lag: The adjustment process is blind and it is difficult to find the optimal solution quickly. Often, the unreasonable formula is only discovered when problems such as hole wall collapse, drill cuttings accumulation and excessive pullback force occur, resulting in project delays and economic losses.

[0009] (3) Lack of global perspective: It is impossible to know the position of the current formula in the entire possible parameter space, and how much "safety margin" it is from the performance degradation boundary.

[0010] Therefore, the HDD industry urgently needs a method that can scientifically, quantitatively, and quickly predict mud performance and globally optimize formulations to reduce construction risks and improve project quality and efficiency. Summary of the Invention

[0011] This invention aims to overcome the aforementioned shortcomings of existing HDD mud preparation technologies and provides a method for predicting and optimizing mud performance based on regression modeling and Monte Carlo simulation. This method can transform limited experimental data into a global formulation-performance graph, intuitively displaying the performance levels achievable by different formulations, thus providing accurate and reliable mud formulation decision support for HDD construction.

[0012] To achieve the above objectives, the present invention adopts the following technical solution:

[0013] A method for predicting and optimizing mud formulation performance for horizontal directional drilling includes the following steps:

[0014] S1: Slurry experiments and data acquisition based on HDD process;

[0015] The geological characteristics and process requirements of HDD (High-Density Digging) construction were analyzed to identify key additives affecting mud performance as independent variables. These mainly included bentonite content (X1), carboxymethyl cellulose (CMC) content (X2), and soda ash content (X3). Experiments covering reasonable value ranges of these additives were designed, and the mud viscosity (Y) corresponding to each formulation was measured using a Marshall funnel viscometer to establish an initial database.

[0016] S2: Establish a nonlinear regression model for HDD mud viscosity;

[0017] Based on the experimental data from step S1, a quantitative prediction model for the relationship between mud viscosity Y and independent variables X1, X2, and X3 is constructed using multivariate nonlinear regression analysis. The model fully considers the interaction effects between components, and its general formula is:

[0018] ;

[0019] The coefficients β are obtained by fitting, and the significance and reliability of the model are ensured by statistical testing.

[0020] S3: Defines the mud performance grade for HDD construction;

[0021] Breaking away from the traditional approach of focusing solely on specific viscosity values, this method classifies mud into multiple performance grades based on viscosity, with each grade corresponding to different construction capabilities and risk levels in HDD (High-Density Digging).

[0022] Level 1 is a low-risk zone: viscosity <50s. It is suitable for pilot holes in soft clay layers that are very easy to drill, but the wall protection ability is weak, and it is prone to hole collapse in complex strata.

[0023] Grades 2-3 are the basic operating range: viscosity 50-60s. Suitable for pilot hole operations under general geological conditions.

[0024] Grades 4-6 represent the optimal pore-expanding zone: viscosity 60-75s. This zone exhibits excellent slag-carrying capacity and wall-protecting properties, and represents the target viscosity range for most pre-pore-expanding stages.

[0025] Grades 7-8 are high-viscosity pullback zones: viscosity ≥ 75s. They possess excellent suspension and lubrication capabilities, making them suitable for large-diameter, long-distance pullback operations or operations in formations prone to borehole collapse.

[0026] S4: Monte Carlo simulation and global performance scan of the formulation space;

[0027] Determine the actual usable range of additives such as bentonite and CMC in HDD field (e.g., X1: 4%-15%, X2: 0.1%-1.5%). Within this range, generate a large number (≥10) of samples using computer processing. 7 (Level) Randomized formulation combinations. Using the regression model established in step S2, the viscosity value of each virtual formulation is quickly predicted.

[0028] S5: Performance grade map of generated HDD mud formulation;

[0029] Visualize the results of step S4:

[0030] Draw a 3D graph with the core components (X1, X2, and X3) as coordinate axes.

[0031] Each random formulation point is assigned a specific region based on the grade defined in S3 to which its predicted viscosity belongs, forming a performance grade distribution cloud map in the full parameter space.

[0032] Isoviscosity isosurfaces are overlaid on the cloud map to clearly mark the boundaries of each level region.

[0033] S6: HDD on-site formulation optimization and decision-making;

[0034] Based on the current geological conditions and construction stage, the construction engineer, from the aforementioned atlas:

[0035] Quick location: Directly locate the formulation area corresponding to the target performance level (e.g., select "Level V" in the pre-expansion stage).

[0036] Optimization selection: Within this region, considering both cost (selecting points with lower additive dosages) and robustness (selecting points far from grade boundaries for greater tolerance), the final optimal formulation is determined.

[0037] Compared with existing technologies, this invention brings the following outstanding advantages to the HDD industry:

[0038] 1. Scientific and standardized: Transforming the "craft" that relies on personal experience into a data-driven scientific method, providing a standardized process and decision-making basis for mud preparation.

[0039] 2. Global Visualization Decision Making: The pioneering "Performance Level Distribution Cloud Map" allows engineers to see the performance results of all possible formulations at a glance, completely eliminating blind trial and error.

[0040] 3. Risk prevention and control: By knowing the performance level of different formulations before construction, accidents in the borehole caused by improper formulations can be proactively avoided, greatly improving construction safety.

[0041] 4. Cost reduction and efficiency improvement: It can quickly identify the lowest cost formula that meets the requirements, saving material costs; at the same time, it reduces on-site commissioning time and speeds up the project progress. Attached Figure Description

[0042] Figure 1 The three-dimensional isosurface wireframe diagram of mud viscosity provided in this embodiment of the invention shows the boundary surfaces between different grades, which can intuitively demonstrate the relationship between mud proportions and performance grades. The black wireframe surfaces represent the boundary surfaces between grades, and the coordinate axes are bentonite content (X1), CMC content (X2), and soda ash content (X3). This diagram is a direct tool for HDD engineers to make formulation decisions.

[0043] Figure 2 The mud viscosity grayscale slices and contour maps provided in the embodiments of the present invention. Figure 2 (a) is a slice with X1 = 9.9% fixed; Figure 2 (b) is a slice with X2 = 0.6% fixed; Figure 2 (c) is a slice with X3 = 0.5% fixed; Figure 2 (d) is a two-dimensional contour plot of a viscosity grade with a fixed X3=0.5%. Detailed Implementation

[0044] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0045] A method for predicting and optimizing mud formulation performance for horizontal directional drilling includes the following steps:

[0046] S1: Slurry experiments and data acquisition based on HDD process;

[0047] The geological characteristics and process requirements of HDD (High-Density Die) construction were analyzed to identify key additives affecting mud performance as independent variables, mainly including bentonite content (X1), carboxymethyl cellulose (CMC) content (X2), and soda ash content (X3). Experiments covering reasonable value ranges were designed, and the mud viscosity (Y) corresponding to each formulation was measured using a Marshall funnel viscometer to establish an initial database.

[0048] S2: Establish a nonlinear regression model for HDD mud viscosity;

[0049] Based on the experimental data from step S1, a quantitative prediction model for the relationship between mud viscosity Y and independent variables X1, X2, and X3 is constructed using multivariate nonlinear regression analysis. The model fully considers the interaction effects between components, and its general formula is:

[0050]

[0051] The coefficients (β) are obtained by fitting the model, and the significance and reliability of the model are ensured by statistical testing.

[0052] S3: Defines the mud performance grade for HDD construction;

[0053] Breaking away from the traditional approach of focusing solely on specific viscosity values, this method classifies mud into multiple performance grades based on viscosity, with each grade corresponding to different construction capabilities and risk levels in HDD (High-Density Digging).

[0054] Level 1 (Low Risk Zone): Viscosity <50s. Suitable for pilot holes in soft clay layers that are very easy to drill, but with weak wall protection and prone to hole collapse in complex strata.

[0055] Grade 2-3 (Basic Operation Area): Viscosity 50-60s. Suitable for pilot hole operations under general geological conditions.

[0056] Grade 4-6 (Optimal pore-expanding zone): Viscosity 60-75s. It possesses good slag-carrying capacity and wall-protecting properties, and represents the target viscosity range for most pre-pore-expanding stages.

[0057] Grade 7-8 (High Viscosity Pullback Zone): Viscosity ≥ 75s. It possesses excellent suspension and lubrication capabilities, suitable for large-diameter, long-distance pullback operations or operations in formations prone to borehole collapse.

[0058] S4: Monte Carlo simulation and global performance scan of the formulation space;

[0059] Determine the actual usable range of additives such as bentonite and CMC in HDD field (e.g., X1: 4%-15%, X2: 0.1%-1.5%). Within this range, generate a large number (≥10) of samples using computer processing. 7 (Level) Randomized formulation combinations. Using the regression model established in step S2, the viscosity value of each virtual formulation is quickly predicted.

[0060] S5: Performance grade map of generated HDD mud formulation;

[0061] Visualize the results of step S4:

[0062] Draw a 3D graph with the core components (X1, X2, and X3) as coordinate axes.

[0063] Each random formulation point is assigned a specific region based on the grade defined in S3 to which its predicted viscosity belongs, forming a performance grade distribution cloud map in the full parameter space.

[0064] Isoviscosity isosurfaces are overlaid on the cloud map to clearly mark the boundaries of each level region.

[0065] S6: HDD on-site formulation optimization and decision-making;

[0066] Based on the current geological conditions and construction stage, the construction engineer, from the aforementioned atlas:

[0067] Quick location: Directly locate the formulation area corresponding to the target performance level (e.g., select "Level V" in the pre-expansion stage).

[0068] Optimization selection: Within this region, considering both cost (selecting points with lower additive dosages) and robustness (selecting points far from grade boundaries for greater tolerance), the final optimal formulation is determined.

[0069] One embodiment of the present invention pertains to a water-based drilling mud. The key components and their ranges are: bentonite (6%–14%), CMC (0.2%–1.0%), and soda ash (0.3%–0.7%). The performance index is the Marvel funnel viscosity (in seconds) measured using a Marvel funnel.

[0070] S1: A three-factor, five-level experimental design was used, and a total of 25 experiments were conducted to obtain the raw data shown in Table 1.

[0071]

[0072] S2: Based on the experimental data obtained in step S1, a multivariate nonlinear regression model is established for the viscosity performance index Y, which is of concern in HDD construction. The model aims to establish the mathematical relationship between the dependent variable (performance index) and multiple independent variables (mud components) and their interaction terms and higher-order terms. Its general form is as follows:

[0073]

[0074] The least squares method is used for model fitting and regression coefficient estimation. The principle of this mathematical method is to find a set of regression coefficient estimates that minimizes the sum of squared residual errors (SSE) between the model predictions and the actual observed values. Its mathematical expression is as follows:

[0075]

[0076] By solving the above optimization problem, the optimal estimates of the regression coefficients can be obtained, thereby determining the final equation.

[0077]

[0078] The model was tested and found to have a coefficient of determination R² > 0.99, indicating that the model has a high goodness of fit and reliable predictive ability.

[0079] S3: Determine the practical usable range of bentonite, CMC, and soda ash additives in HDD practice (e.g., X1: 4%-15%, X2: 0.1%-1.5%). Within this range, generate a large number (≥10) of samples using computer processing. 7 (Level) Randomized formulation combinations. Using the regression model established in step S2, the viscosity value of each virtual formulation is quickly predicted, as shown below. Figure 1 The three-dimensional wireframe contour map shown can be sliced ​​to generate a clearer two-dimensional contour map.

[0080] S4: If engineers need to prepare mud with a viscosity between 60-65s, they only need to... Figure 1 We found a spatial region between the 60s and 65s isosurfaces, where all combinations of (X1, X2, X3) are potential formulations, enabling rapid screening and optimization of formulations.

[0081] S5: Simultaneously, a grayscale contour map of the slice at a soda ash content of 0.5% was generated. Figure 2 This diagram can provide more precise guidance on how to adjust the amounts of bentonite and CMC to achieve the target viscosity when the amount of soda ash is fixed.

Claims

1. A method for predicting and optimizing the performance of mud formulations for horizontal directional drilling, characterized in that, Includes the following steps: S1: Slurry experiments and data acquisition based on HDD process; Analyzing the geological features and technological requirements of HDD (High-Density Digging) construction, key additives affecting mud performance were identified as independent variables, including bentonite content. X 1. Carboxymethyl cellulose (CMC) content X 2. Soda ash content X 3. Use a Marsh funnel viscometer to measure the slurry viscosity corresponding to each formulation. Y Establish the original database; S2: Establish a nonlinear regression model for HDD mud viscosity; Based on the experimental data from the original database in step S1, multivariate nonlinear regression analysis was used to construct the relationship between mud viscosity Y and independent variables. X 1, X 2, X A quantitative prediction model for the interaction between the three components; the model fully considers the interaction effects between the components, and its general formula is: ； The coefficients of each term are obtained through fitting. β And statistical tests are used to ensure the significance and reliability of the model; S3: Defines the mud performance grade for HDD construction; Based on viscosity values, the mud is classified into multiple performance grades, each grade corresponding to different construction capabilities and risk levels of HDD; S4: Monte Carlo simulation and global performance scan of the formulation space; Determine the actual usability range of bentonite and CMC additives in the HDD field; within this actual usability range, generate random formulation combinations using a computer; use The regression model established in step S2 can quickly predict the viscosity value of each virtual recipe. S5: Performance grade map of generated HDD mud formulation; Visualize the results of step S4: With core ingredients X 1, X 2 and X 3. Use coordinate axes to plot a 3D graph; Each random formulation point is assigned a specific region based on the grade defined in S3 to which its predicted viscosity belongs, forming a performance grade distribution cloud map in the full parameter space; Isovisible isosurfaces are overlaid on the cloud map to clearly mark the boundaries of each level region; S6: HDD on-site formulation optimization and decision-making; Based on the current geological conditions and construction stage, the following information is derived from the HDD mud formulation performance grade map: Quick location: Directly locate the formula area corresponding to the target performance level; Optimization selection: Taking into account both cost and robustness within the region, determine the final optimal formulation.

2. The method for predicting and optimizing mud formulation performance for horizontal directional drilling according to claim 1, characterized in that, In S3, level 1 is a low-risk zone: viscosity <50s; suitable for pilot holes in soft clay layers that are very easy to drill, but with weak wall protection and prone to hole collapse in complex strata; Grades 2-3 are the basic operating range: viscosity 50-60s; Suitable for pilot hole operations under general geological conditions; Grades 4-6 represent the optimal pore-expanding zone: viscosity 60-75s; it possesses excellent slag-carrying capacity and wall-protecting properties, and is the target viscosity range for most pre-pore-expanding stages; Grades 7-8 are high-viscosity pullback zones: viscosity ≥ 75s; they have excellent suspension and lubrication capabilities and are suitable for large-diameter, long-distance pullbacks or operations in formations prone to borehole collapse.

Images