Aviation kerosene uses jet fuel No.3 (JET A⁃1), which is a common source of danger in leaking fires.Based on theoretical analysis and PHAST software simulation, the influence of a tank leak in the aviation coal tank area on the surrounding storage tank was studied. The method for determining the safety distance between the tanks was introduced. The results show that the thermal radiation intensity increases with the increase of wind speed, and the thermal radiation value at the maximum wind speed increases by 65.7% compared with the average wind speed. As the distance increases, the radiation value becomes smaller, and as the height of the tank increases, the heat radiation value becomes larger, and the atmospheric humidity has less influence on the heat radiation. It is determined that the safety distance of the oil tank of the project is 17 m. The theoretical calculation is carried out by using MATLAB software and compared with the numerical simulation results. The maximum error is 18.7%, which has strong reliability. The research results have important theoretical and practical significance for tank tank design and guiding emergency disaster relief.

In order to explore the new mode of using residual oil, Ansai oilfield first used horizontal wells to encrypt the remaining oil, and 7 key technologies and field tests were carried out. Based on the different way of encryption horizontal well spacing and the transformation parameter study, the mechanism of remaining oil distribution was obtained and the technology of using remaining oil in high and medium water-bearing encrypted horizontal well adapted to Ansai low permeability reservoir was explored. It is proved that horizontal encryption can effectively use the remaining oil and low abundance reservoir between wells to utilize the remaining reserves of developed areas, which provides new enlightenment for the same reservoir to further improve the recovery rate.

With the implementation of the three oil recovery technology, although polymer flooding technology improves the oil recovery, it also brings new challenges to crude oil processing and gathering. High return pressure of oil gathering pipeline was caused by wax deposition of polymer flooding pipeline. Study was carried out on the heat cleaning of the polymer flooding gathering pipeline between the wellhead and the metering room after shutdown the well. The effects of heat washing water quantity, hot wash water temperature and different concentration of polymer on the dredging time were studied under different oil gathering radius. It was found that the hot wash water temperature was more obvious in the range of 65~70 ℃, which was calculated by the theoretical formula. When cleaning different polymer concentration pipelines, in the low water conditions, cleaning time of produced liquid with high concentration polymer was relatively longer. With the increase of heat washing water, there was no obvious change in the cleaning time of the sediment with different concentration.

Source rocks evaluation is of great importance for gas exploration in Changling depression of southern Songliao basin. Aimed to make clear the TOC distribution of source rocks in Shahezi formation, based on the logging data and core tested data, variable coefficient ΔlgR model was used to calculate the TOC quantitatively, source rocks of different TOC grade and their thickness distribution were described. The key coefficients in the model were calculated using variable coefficient ΔlgR model according to the sampled data of the studied area, and TOC calculation results can reflect the TOC underground accurately. TOC predicting error by variable coefficient ΔlgR model was about 19.1%, which was reduced 17.5% compared to traditional ΔlgR model. Source rocks in Changling depression are characterized by high TOC value, middle less source rocks, good source rocks and excellent source rocks account for 26.1%, 18.2% and 55.7% of all the source rocks in volume, respectively. Good and excellent source rocks mainly distributed in Heidimiao sag, the maximum thickness is over 400 meters, Haerjin and Qianbei sags also developed good source rocks the general thickness is 200~300 meters.

The properties of DongfangYangpuHaikou and DongfangSanya pipeline natural gas were analyzed, the content of CO2 was 4.94%(mol) and 8.21%(mol) respectively, which was more than 3%(mol) required in standard (Q/SY 30—2002). The content of H2S was 25.2 mg/m3 in the DongfangSanya pipeline natural gas, which was also more than standard of 20 mg/m3. So, the chemical corrosion of DongfangSanya pipeline was serious. By the test of sulfate reducing bacteria cultivating, sulfate reducing bacteria corrosion was proved in DongfangYangpuHaikou pipeline, although it was very slight.

Using three-dimensional model, the cross-shaped baffler distributor was systematically simulated by CFX fluid mechanics modeling software. The impact of size of upon cover of inner sleeve on degree of uniformity of gas distribution of feeding distributor was emphasized. Meantime, the impact of the number and size of guide plate on performance of entrainment and pressure drop of the distributor was researched. At last, the structural parameters of excellent performanced cross-shaped baffler distributor was confirmed and so as to support direction for structural optimization of the distributor.

A hybrid modeling method for the complex process was proposed．The process，which was highly nonlinearity，had interacting independent variables， and had process data belonged to different categories． This method combined self－organizing map network （SOM network），principal component analysis （PCA），and relevance vector machine （RVM）．First，
divided the sample datum into several subspaces which had the same patterns with the utilizing of SOM，the pattern space separation was realized．Then derived the principal components based on the modeling samples of each subspace，defined the optimal principal component numbers based on prediction ability，and filtered the redundancy．At last，the principal component
of each subspace was applied as the input of RVM，building the models individually ，and the classifying modeling was realized based on the separated of sample pattern spaces．The simulation results and the application in 4－carboxybenzaldehyde content soft sensor of pureed terephthalic acid production show that the accuracies of regressing and predicting of SOM－PCA －RVM model is better than both RVM model and PCA － RVM model．