Train A
Train B
Best MRA Fit
Best MRA Fit
R2 =
0.869769
R2 =
0.840843
Train A Salt out, Ptb = INT+(Flow, m/Hr *C1)+(Flow, m/hr^2 *C2)+ salter Train 15A, Amperage *C9)+(Desalter Train 15A, A *C15)+(15 A Mix valve DP at Desalter, bar^2 *C16)+(16 Flow, mh Flow, m/h Desalter T Desalter T
Train A Salt out, Ptb = INT+(Flow, m/Hr *C1)+(Flow, m/hr^2 *C2)+ salter Train 15 B , Amperage *C9)+(Desalter Train 15A, A *C15)+(15 B Mix valve DP at Desalter, bar^2 *C16)+(16 Flow, mh Flow, m/h Desalter T Desalter T
Wash water
C1
C2
C3
C4
C1
C2
C3
C4
INT
INT
0.007013 0.003513 1.996093
0.001600 0.003900 0.410800
-2.06941 1.048238 -1.97418
-0.4446 1.179456 -0.37695
-0.000139 4.95E-05 2.817826
-3.8E-05 6.72E-05 -0 . 56834
-0.1867 0.066472 -2.80868
0.051612 0.090172 0.57237
-128.298 401.0411 -0.31991
-16.5092 30.22281 -0.54625
Coef. Std Er T-Value
Coef. Std Er T-Value
Mix valve DP based on Crude ow
SSR
SSE
MSR MSE
F
S^2
SSR
SSE
MSR MSE
F
S^2
2.35137 0.086802 0.029764 2.916314 0.172523
2.891859 0.094998 0.036606 2.595167 0.191300
2.08324
2.279957
Independent variable rank from linear regression
Independent variable rank from linear regression
T-value
Variable name
T-value
Variable name
2.008296 1.448384 1.195132 1.019147 0.807357 0.755804 0.529131 0.507686 0.136056 0.128649 0.12392 0.080569 0.049237
10 16 A Mix valve DP at Desalter, bar
2.425747 2.21572 2.142608 2.042593 1.394086 1.219738 1.194885 1.164472 1.055314 0.96255 0.759978 0.536674 0.326651
8 Desalter train 16B, Amperage 10 16B Mix valve DP at Desalter, bar 5 Desalter train 15B, Voltage 7 Desalter train 16A, Voltage
2 Crude ow, m/hr
9 15 A Mix valve DP at Desalter, bar 8 Desalter train 16A, Amperage 6 Desalter train 15A, Amperage
2 Crude ow, m/hr 12 15B Wash water %
1 INT
9 15B Mix valve DP at Desalter, bar 11 Train B: Demulsier dosage, ppm
7 Desalter train 16A, Voltage
12 15A Wash water %
13 16B wasg water % 3 Desalter temp 15B, C 4 Desalter temp 16B, C
11 Train A: Demulsier dosage, ppm
3 Desalter temp 15A, C 4 Desalter temp 16A, C 13 16A wash water %
6 desalter train 15B, Amperage
5 Desalter train 15A, Voltage
1 INT
Figure 5 MRA output
More importantly, when solids in crudes were high, a 5-10% measurement difference was observed due to solids stabili- sation at the interface phase. The measurement discrepancy was resolved with the commencement of solids wetting aids. v Wash water: The refinery operates with the same crude and process conditions for its Trains A and B. Desalter Train B exhibited a higher desalted crude salt outlet compared to Train A. Data analysis revealed that the wash water flow meter valve actuator was malfunctioning, causing discrep- ancies in wash water flow rates to Train B. After tuning, the wash water flow rates were normalised, addressing the difference in salt removal efficacy between the trains (see Figure 6 ). w Slop oil: Effective slop oil management is critical for con- sistent desalter performance. The survey identified that slop oil handling facilities are always undersized, leading to major challenges with respect to segregation and pretreatment, which results in inconsistent slop oil volume reprocessing.
Optimisation protocols limited slop reprocessing to less than 3% of crude flow for white slop and less than 1% for effluent treatment plant (ETP) recovered slop. Any slop processing above these limits requires pretreatment and analysis. x Desalter brine pH: Water-soluble acidic impurities in crudes sourced via upstream additives result in a drop in desalter brine pH to 3-4, despite wash water pH ~7-8. The acidic impurities from the crude source were confirmed with crude tank water drain pH ~2.7-3.0. The acidic brine not only raises concerns about high corrosion in the desalter but also leads to sulphide carryover in the desalter brine, yield- ing major environmental, health, and safety (EHS) concerns in the wastewater treatment plant. Additionally, an increase in chloride yield in overheads was observed, even with the same salt level in desalted crude. Both literature searches and validation experiments in the lab revealed that the rea- sons for the increase in chloride in overheads are due to acid impurities catalysing salt hydrolysis at high-temperature zones, such as crude heaters and the flash zone. To control the salt hydrolysis reactions from acid impurities, an organic- based pH modifier, PROCHEM, was deployed to increase the brine pH to 5.5-6.5. Note: It is crucial to control brine pH below 7.0 with organic pH modifiers, as excessive dos - ing of additives to pH above 7.0 will lead to partitioning of impurities in the crude, which can increase salt potential in overheads. The lack of reliable online pH monitoring and control restricts the use of a strong base (caustic soda) to control pH. Hence, organic weak-based pH control is pre- ferred to reduce impacts downstream, including sodium limits in atmospheric residue cuts. y Mud wash: During stable emulsion and thick rag layer formation in the desalter, the analysis of the rag layer reveals high solids, leading to pickering emulsion. Although the introduction of solids wetting agents improved solids removal and prevented rag layer build-up, solids removal
Figure 6 Wash water flow to mix valve – flow discrepancies – actuator fluctuations
94
PTQ Q4 2024
www.digitalrefining.com
Powered by FlippingBook