FILTRATION METHOD EFFICIENTLY DESALTS CRUDE IN COMMERCIAL TEST

During 3 months of industrial testing of a filtration crude oil desalting method, a total of 120,500 metric tons (mt), or 1,475 mt/d (almost 11,000 b/d) of crude was processed. Rongxi Du, Kai Peng, and Li Wang, engineers at Wuhan Petrochemical Works, Wuhan, China, in an unpublished report, indicate that they determined unit operating parameters and performed statistical analyses of desalting-efficiency data from the test run.

May 17, 1993

4 min read

During 3 months of industrial testing of a filtration crude oil desalting method, a total of 120,500 metric tons (mt), or 1,475 mt/d (almost 11,000 b/d) of crude was processed.

Rongxi Du, Kai Peng, and Li Wang, engineers at Wuhan Petrochemical Works, Wuhan, China, in an unpublished report, indicate that they determined unit operating parameters and performed statistical analyses of desalting-efficiency data from the test run.

The engineers also determined relationships between desalting efficiency and flow velocity, relative density, mixing pressure drop (MPD), filtration-tank pressure drop, and temperature.

The desalting and dewatering level of single-stage filtration desalting was found to be equal to that of two-stage electrostatic desalting with remarkable benefits resulting from reduced power, water, and demulsifier requirements.

FILTRATION DESALTING

Fig. 1 shows the desalting process.

A single-stage filtration desalting unit (FDU) was run in parallel with an electrostatic desalting unit (EDU). During the testing in late 1991, about 25% of the total EDU processing volume was run through the FDU.

A highly hydrophilic, fairly lipophilic filter material is filled in the filtration tank. When emulsified crude oil enters the tank, absorption and fluid forces act on it.

The crude undergoes absorption, collision, moistening, and filtration. Water and oil eventually separate in the settling tank.

The inlet valve of the first electrostatic desalting tank (EDT-1) is turned down to increase the pressure drop of the EDU. The crude oil is thus divided into two routes which join and enter EDT-3 (Fig. 1).

TEST PARAMETERS

Table 1 lists oil types, relative densities, and salt contents of the feedstocks processed.

Yizheng's machinery impurity and emulsified water contents are 5.0% and 20%, respectively, while the Yizheng/Oman mixture's oil machinery impurity and emulsified water are 3.0% and 16%.

RESULTS

Table 2 compares desalting efficiencies for the EDU and FDU. Du, Peng, and Wang concluded the use of an improper demulsifier caused the unsteady desalting rate and high water content in Period 3.

Higher temperatures were found to increase FDU desalting efficiency. The engineers also determined that injecting more water created a stronger emulsion. Additional water also decreased the temperature, thus worsening the separation efficiency.

From the statistical analysis of the data, Du, Peng, and Wang concluded that the factors affecting desalting efficiency, from most to least important, are velocity, relative density, mixing pressure drop, resistance pressure drop (RPD), and temperature (Fig. 2).

REVIVICATION

In the first two of the six test periods, the filter was so active that RPD increased slowly. But during Periods 3-5, incomplete reverse watering and filter revivification decreased the cycle times and caused mechanical impurities to accumulate in the filter.

These impurities caused demulsification activity to decrease and tank pressure drop to increase sharply (Table 3). Operation had to be stopped to revivify the filter.

Before the fifth test period, the revivification process was improved and desalting ability returned to initial levels. After the sixth period, further improvements resulted in RPD increasing only 0.03 MPa, indicating the filter was in the best condition yet observed.

During Periods 1-3, reverse watering was accomplished using potential heat and the remaining pressure of the oil, which acted together to push the water. Results were improved by using a water pump from Period 4 forward. Water velocity was 80 tons/hr.

Table 4 shows the power consumption comparison of the EDU and FDU. The EDU data are from 1990. The Table clearly shows greatly decreased power consumption with the FDU.

Water consumption for EDT-1 and 2 were 7% in 1990, compared with 3-4% for the FDU. Filtration desalting also saves demulsifier, says the research team.

Du, Peng, and Wang say the process 2 is suitable for all oils with d4 20 between 0.8500 and 0.915. The desalting rate and water content of the product are equivalent to two-stage EDU levels.

The process is easy to operate, and reverse watering is practical and convenient.