Research Article

Prediction of the Control Effect of Fractured Leakage in Unconventional Reservoirs Using Machine Learning Method

Table 2

Statistical analysis and summary of sample data.

TypesMinimumMaximumRangeArithmetic meanStd. deviationSkewnessKurtosis

Leak rate (m3/h)31009 751.3769233.572440.430229-1.35702
Displacement (m3/h)30100.87 0.852.6615418.324241.1268491.879827
Pumping pressure max (MPa)5161 110.730772.739423-0.021660.004141
Concentration (%)0.050.36360.31360.1097920.0570743.44208515.20886
Addition of KZ-3 (t)0220.4615380.7956981.3504040.010491
Addition of KZ-4 (t)02.52.50.9423080.858520.238694-1.44599
Addition of KZ-5 (t)02.52.50.1346150.5106124.2425918.6555
Addition of walnut shell (t)0331.9230770.873678-0.57641-0.13565
Amount of vermiculite (t)0330.6346150.9152311.1733390.043976
Amount of comprehensive plugging agent (t)0331.4230771.080352-0.17773-1.36465
Addition amount of LWD agent (t)0220.1346150.4714483.50058711.44695
Total addition (m3)11806955.4230813.31042-1.020633.95618
Density (g/cm3)1.151.250.11.2215380.022819-2.326475.916531
Funnel viscosity (s)46722658.038469.5009340.36903-1.50663
Actual blocking effect (%)0.110.90.5969230.3301020.280503-1.75597