1. Water Management in Process Plants
David Puckett
Débora Campos de Faria
Miguel J. Bagajewicz

2. Sources of Refinery Wastewater
Caustic Treating
NH3 and H2S Water Contamination
Distillation
Water Contamination with Organics
Amine Sweetening
NH3 and H2S Water Contamination
Merox Sweetening
NH3 and H2S Water Contamination
Water Contamination with NH3, H2S, and Organics
Hydrotreating
Desalting
Saline Water Contamination

3. Water Management Methods
Wastewater produced in industrial processes can be handled in three fashions.
End-of-Pipe Cleanup
Reuse
Regeneration

4. Regeneration Methods
API separator and activated carbon to remove organics from distillation and hydrotreating wastewater.
Reverse osmosis to remove saline contamination from desalting wastewater.
Chevron wastewater treatment to remove acid gas contamination from caustic treating, sweetening, and hydrotreating wastewater.

5. Wastewater Optimization
Current methods of optimizing water reuse and regeneration rely on several assumptions.
Operating and capital costs are functions solely of treated water flow rate.
Fixed process outlet concentrations.

6. Wastewater Optimization
FCI
$458000
22% of the amount of contaminant removed,
89% of the FCI B
FCI
$408000
51% of the amount of contaminant removed,
78% of the FCI C
FCI
$317000

7. Wastewater Optimization
Depending on the contaminants present and the treatment processes used, the assumption that regeneration costs are dependent solely on flow rate may not be valid.
The optimum solution for a water allocation problem must take into account factors other than flow rate.

8. Removal of Organics Distillation Wastewater Contaminated with Organics
Hydrotreating
API Separator and Activated Carbon Adsorber
Wastewater Free of Organics

9. API Separator
Removes multi-phase contamination through differences in specific gravity.

10. API Separator
Appropriate for use with any contaminant that forms a distinct phase in the process water.
Oil and Light Organics
Organic and Inorganic
Sediment

11. API Separator Simulation
The basis of the separation is Stokes’ Law.
For a given contaminant in water, rate of settling is determined solely by contaminant particle size.
Quality of separation can be improved through flocculation and coagulation.
Buoyant Force
Drag Force
Gravitational Force

12. API Separator Simulation
Process water contaminant concentration does not change quality of separation.
Percentage of contaminants removed on a volume basis determined based on a normal distribution of particle radii.

13. Quality of Separation vs. Length
Separator Depth = 1 m
Separator Width = 2 m
Entrance to Separator at 0.5 m
Process Water Flow Rate = 1 m3 / s
Contaminant SG = 0.95
Mean Contaminant Diameter = 0.5 mm
Quality of separation improves with increasing length, but with diminishing returns.

14. Quality of Separation vs. Specific Gravity
Separator Depth = 1 m
Separator Width = 2 m
Separator Length = 25 m
Entrance to Separator at 0.5 m
Process Water Flow Rate = 1 m3 / s
Mean Contaminant Diameter = 0.5 mm
Separation quality is poor for contaminants similar in density to water.

15. Quality of Separation vs. Particle Diameter
Separator Depth = 1 m
Separator Width = 2 m
Separator Length = 25 m
Entrance to Separator at 0.5 m
Process Water Flow Rate = 1 m3 / s
Contaminant SG = 0.95
Quality of separation improves with increasing particle diameter, but with diminishing returns.

16. Quality of Separation vs. Wastewater Velocity
Separator Depth = 1 m
Separator Width = 2 m
Separator Length = 25 m
Entrance to Separator at 0.5 m
Contaminant SG = 0.95
Mean Contaminant Diameter = 0.5mm
Quality of separation improves with decreasing velocity. A velocity of zero would give perfect separation.

17. Quality of Separation vs. Settling Distance
Separator Depth = 1 m
Separator Width = 2 m
Separator Length = 25 m
Contaminant SG = 0.95
Process Water Flow Rate = 1 m3 / s
Quality of separation improves with decreasing settling distance. Separators that handle oil will have entrance as close to water surface as possible.

18. Equipment Cost vs. Flow rate
Separator Depth = 1 m
Separator Width = 2 m
Mean Particle Diameter = 1 mm
Process Water Flow Rate = 0.5 m3 / s
Equipment cost is sensitive to both flow rate and separation quality.

19. Operating Cost vs. Flow rate
Separator Depth = 1 m
Separator Width = 2 m
Mean Particle Diameter = 1 mm
Process Water Flow Rate = 0.5 m3 / s
Operating cost is sensitive only to flow rate.

20. API Separator Performance
With a normal distribution of particle diameters, quality of separation can be solved analytically.
A bit impractical to implement.

21. API Separator Performance
Varying SG
Varying h
Varying L
Varying Dp
Varying F/A
The approximation is quite close for all variables. The worst fit is for changes in settling height.

22. API Separator Equipment Cost
Varying ΔSG, Dp, h
Varying F and %QS
Equipment cost is dependent on flow rate, quality of separation, specific gravity of contaminant, contaminant particle size, settling distance, and the price of steel.

23. API Separator Operating Cost
Operating cost is dependent solely on flow rate.

24. Activated Carbon
Removes soluble contaminants through adsorption onto the activated carbon surface.

25. Activated Carbon
Appropriate for use with liquid or gaseous contaminants that are water soluble or form emulsions.
Dissolved Organics
Insoluble Organics of
< 150 microns
Droplet Size
Dissolved Gases

26. Activated Carbon Simulation
Separation will follow the Langmuir isotherm.
For the Langmuir isotherm, the rate of adsorption, assuming negligible pore holdup and spherical adsorbate particles, is as follows.

27. Activated Carbon Simulation
For a fixed bed adsorber,
process water should reach
equilibrium with activated
carbon prior to end of bed.
A constant length of bed
is required for adsorption.

28. Water Treatment Rate vs. Adsorbant Surface Area
Contaminant Diffusivity = 5 * 10 ^ -11 m2 / s
Langmuir Coefficient = 0.06 m3 / kg
Saturation Adsorption = 0.4 kg / kg adsorbant
Inlet Contaminant Concentration = 0.25 kg / m3
Greater adsorbant surface area results in faster adsorption and a faster rate of treatment.

29. Time in Service vs. Adsorber Diameter
Contaminant Diffusivity = 5 * 10 ^ -11 m2 / s
Langmuir Coefficient = 0.06 m3 / kg
Saturation Adsorption = 0.4 kg / kg adsorbant
Inlet Contaminant Concentration = 0.25 kg / m3
A greater diameter has more adsorbant per unit length and thus will take longer to saturate.

30. Time in Service vs. Adsorber Height
Contaminant Diffusivity = 5 * 10 ^ -11 m2 / s
Langmuir Coefficient = 0.06 m3 / kg
Saturation Adsorption = 0.4 kg / kg adsorbant
Inlet Contaminant Concentration = 0.25 kg / m3
The saturation wave travels through the adsorber at a constant speed.

31. Outlet Concentration vs. Inlet Concentration
Contaminant Diffusivity = 5 * 10 ^ -11 m2 / s
Langmuir Coefficient = 0.06 m3 / kg
Saturation Adsorption = 0.4 kg / kg adsorbant
Outlet concentration from adsorber is dictated by adsorption thermodynamics.

32. Equipment Cost vs. Flow Rate
Contaminant Diffusivity = 5 * 10 ^ -11 m2 / s
Langmuir Coefficient = 0.06 m3 / kg
Saturation Adsorption = 0.4 kg / kg adsorbant
Inlet Contaminant Concentration = 0.25 kg / m3
Wastewater pump and column diameter must scale for flowrate.

33. Operating Costs vs. Flowrate
Contaminant Diffusivity = 5 * 10 ^ -11 m2 / s
Langmuir Coefficient = 0.06 m3 / kg
Saturation Adsorption = 0.4 kg / kg adsorbant
Inlet Contaminant Concentration = 0.25 kg / m3
A greater flow rate means more regenerations per year in addition to increased pumping work.

34. Equipment Cost vs. Inlet Concentration
Contaminant Diffusivity = 5 * 10 ^ -11 m2 / s
Langmuir Coefficient = 0.06 m3 / kg
Saturation Adsorption = 0.4 kg / kg adsorbant
Inlet Contaminant Flow Rate = m3 / hr
No changes in the adsorber need to be made to accommodate a greater inlet concentration.

35. Activated Carbon Aerogel
Contaminant Diffusivity = 5 * 10 ^ -11 m2 / s
Langmuir Coefficient = 0.06 m3 / kg
Saturation Adsorption = 0.4 kg / kg adsorbant
Inlet Contaminant Flow Rate = m3 / hr
A greater inlet concentration has the same effect as a greater flow rate. More contaminant must be adsorbed necessitating more regenerations.

36. Activated Carbon Performance
Outlet concentration can be calculated analytically.

37. Activated Carbon Equipment Cost
Varying F and CIN
Varying H
Varying D
Equipment cost is dependent on flow rate, inlet concentration, column measurements, and the prices of steel and activated carbon.

38. Activated Carbon Operating Cost
Varying CIN
Varying F
Operating cost is dependent only on flow rate and concentration.

39. Removal of Salts Wastewater Contaminated with Salts Desalting
Reverse Osmosis Separation
Wastewater Free of Salts

40. Reverse Osmosis
Removes salts from process water by forcing water against the salt concentration gradient.

41. Reverse Osmosis Suitable for the removal of any soluble contamination.
Soluble Salts
Soluble Organics
Microorganisms

42. Reverse Osmosis Simulation
Separation proceeds based on Fick’s First Law.
For reasonably dilute solutions, the van’t Hoff approximation of osmotic pressure can be used.
Quality of separation is fixed by type of membrane used.

43. Flow Rate vs. Membrane Area
Membrane Thickness = m
Membrane Permeability = 1.92 * 10-9 ((m3 m) / (m2 sec atm))
Brine Pressure = 10 atm
Brine Ion Concentration = 100 mol / m3
Ion Rejection Percentage = 0.99
Flow rate and membrane area are linearly related, as would be expected from Fick’s Law.

44. Flow Rate vs. Membrane Thickness
Membrane Area = 100 m2
Membrane Permeability = 1.92 * 10-9 ((m3 m) / (m2 sec atm))
Brine Pressure = 10 atm
Brine Ion Concentration = 100 mol / m3
Ion Rejection Percentage = 0.99
Flow rate and membrane thickness are inversely related, as would be expected from Fick’s Law.

45. Flow Rate vs. Brine Pressure
Membrane Area = 100 m2
Membrane Thickness = m
Membrane Permeability = 1.92 * 10-9 ((m3 m) / (m2 sec atm))
Brine Ion Concentration = 100 mol / m3
Ion Rejection Percentage = 0.99
Flow rate is zero when the pressure gradient is equal and opposite to the osmotic pressure gradient.

46. Flow Rate vs. Rejection Percentage
Membrane Area = 100 m2
Membrane Thickness = m
Membrane Permeability = 1.92 * 10-9 ((m3 m) / (m2 sec atm))
Brine Pressure = 10 atm
Brine Ion Concentration = 100 mol / m3
A higher rejection percentage results in a larger osmotic pressure gradient.

47. Flow Rate vs. Brine Concentration
Membrane Area = 100 m2
Membrane Thickness = m
Membrane Permeability = 1.92 * 10-9 ((m3 m) / (m2 sec atm))
Brine Pressure = 10 atm
Ion Rejection Percentage = 0.99
Again, flow rate is zero when the pressure gradient is equal and opposite to the osmotic pressure gradient.

48. Flow Rate vs. Temperature
Membrane Area = 100 m2
Membrane Thickness = m
Membrane Permeability = 1.92 * 10-9 ((m3 m) / (m2 sec atm))
Brine Pressure = 10 atm
Brine Concentration = 100 mol/m3
Ion Rejection Percentage = 0.99
The van’t Hoff approximation introduces a dependence of osmotic pressure on temperature.

49. Equipment Cost vs. Flow Rate at 1463 ppm Inlet
Membrane Thickness = m
Membrane Permeability = 9.17 * ((m3 m) / (m2 sec atm))
Brine Pressure = 10 atm
Base Ion Rejection Percentage = 0.8
Equipment cost increases exponentially with process water purity as membrane rejection is fixed so membranes must be worked in series to achieve higher purity.

50. Equipment Cost vs. Flow Rate at 59 ppm Inlet
Membrane Thickness = m
Membrane Permeability = 9.17 * ((m3 m) / (m2 sec atm))
Brine Pressure = 10 atm
Base Ion Rejection Percentage = 0.8
Equipment costs are dependent on the relative inlet/outlet concentrations, not the absolute concentrations.

51. Operating Cost vs. Flow Rate at 1463 ppm Inlet
Membrane Thickness = m
Membrane Permeability = 9.17 * ((m3 m) / (m2 sec atm))
Brine Pressure = 10 atm
Base Ion Rejection Percentage = 0.8
The same trends as observed in equipment costs are observed in operating costs.

52. Operating Cost vs. Flow Rate at 59 ppm Inlet
Membrane Thickness = m
Membrane Permeability = 9.17 * ((m3 m) / (m2 sec atm))
Brine Pressure = 10 atm
Base Ion Rejection Percentage = 0.8
The same trends as observed in equipment costs are observed in operating costs.

53. Reverse Osmosis Performance
Outlet concentration defined by membrane properties.

54. Reverse Osmosis Flow Rate

55. Reverse Osmosis Equipment Cost
Varying F and CIN
Equipment costs are dependent on flow rate, brine concentration, and membrane cost for single membrane. Cost scales based on bypass ratio and number of membranes in series for a series of membranes.

56. Reverse Osmosis Operating Cost
Varying F
Operating costs are dependent solely on flow rate. Cost scales based on bypass ratio and number of membranes in series for a series of membranes.

57. Removal of H2S and NH3 Caustic Treating
Wastewater Contaminated with H2S and NH3
Amine Sweetening
Merox Sweetening
Hydrotreating
Chevron Wastewater Treatment
Wastewater Free of H2S and NH3

58. Chevron Wastewater Treatment
Removes dissolved gases from wastewater through stripping and absorption.

59. Chevron Wastewater Treatment
Suitable for the removal of any suitably volatile contaminant.
Hydrogen Sulfide
Ammonia

60. Chevron Wastewater Treatment Simulation
Equilibrium Line
m=L/G
b=Yo-XI(L/G)
McCabe-Thiele Method can be used.

61. Quality of Stripping vs. Reboil Ratio
Inlet H2S Concentration = 1000 mol/m3
Tray # = 6
Outlet Gas = 50% H2S
Superior quality of separation achieved with less of the wastewater boiled.

62. Quality of Stripping vs. Number of Trays
Inlet H2S Concentration = 1000 mol/m3
Outlet Gas = 50% H2S
Reboil Ratio = 0.6
Superior separation achieved at greater number of trays, though diminishing returns are noted.

63. Quality of Stripping vs. Inlet Concentration
Inlet H2S Concentration = 1000 mol/m3
Outlet Gas = 50% H2S
Reboil Ratio = 0.6
Tray # = 6
An increase in inlet concentration will always increase the outlet concentration if all other factors remain constant.

64. Equipment Cost vs. Flow Rate
Outlet Gas = 50% H2S
Reboil Ratio = 0.6
Tray # = 6
Equipment costs increase in stepped fashion based on need for additional tray to maintain specified outlet concentrations.

65. Operating Cost vs. Flow Rate
Outlet Gas = 50% H2S
Reboil Ratio = 0.6
Tray # = 6
Operating costs insensitive to inlet concentration as primary operating costs is heating of large amounts of water.

66. Equipment Cost vs. Flow Rate
Inlet H2S Concentration = 1000 mol/m3
Outlet Gas = 50% H2S
Reboil Ratio = 0.6
Same trends as observed with varied inlet concentrations.

67. Operating Cost vs. Flow Rate
Inlet H2S Concentration = 1000 mol/m3
Outlet Gas = 50% H2S
Reboil Ratio = 0.6
Same trends as observed with varied inlet concentrations.

68. Chevron Wastewater Treatment Simulation
(YI,YI)
m=Reflux Ratio
McCabe-Thiele Method can be used.

69. Distillation Quality vs. Number of Trays
Inlet NH3 Concentration = 1000 mol/m3
Outlet Gas = 98% NH3
Reboil Ratio = 0.6
Reflux Ratio = 0.6
Same trends as observed with stripping column.

70. Distillation Quality vs. Reboil Ratio
Inlet NH3 Concentration = 1000 mol/m3
Outlet Gas = 98% NH3
Reflux Ratio = 0.6
Stripping Tray # = 6
Same trends as observed with stripping column.

71. Distillation Quality vs. Reflux Ratio
Inlet NH3 Concentration = 1000 mol/m3
Outlet Gas = 98% NH3
Reboil Ratio = 0.6
Stripping Tray # = 6
Increasing reflux ratio will improve the quality of the separation. Note the break in the graph at the minimum reflux ratio.

72. Equipment Cost vs. Flow Rate
Outlet Gas = 98% NH3
Reboil Ratio = 0.6
Reflux Ratio = 0.6
Stripping Tray # = 6
Same trends as observed with stripping column.

73. Operating Cost vs. Flow Rate
Outlet Gas = 98% NH3
Reboil Ratio = 0.6
Reflux Ratio = 0.6
Stripping Tray # = 6
Same trends as observed with stripping column.

74. Equipment Cost vs. Flow Rate
Inlet NH3 Concentration = 1000 mol / m3
Outlet Gas = 98% NH3
Reboil Ratio = 0.6
Reflux Ratio = 0.6
Same trends as observed with stripping column.

75. Operating Cost vs. Flow Rate
Inlet NH3 Concentration = 1000 mol / m3
Outlet Gas = 98% NH3
Reboil Ratio = 0.6
Reflux Ratio = 0.6
Same trends as observed with stripping column.

76. Performance of Chevron Wastewater Treatment
H2S Removal
NH3 Removal

77. Chevron Wastewater Treatment Equipment Costs
Varying Tray Number
Varying D
Varying F
Equipment costs strongly dependent on number of trays, column diameter, and the price of stainless steel.

78. Chevron Wastewater Treatment Operating Costs
For both the stripping and the distillation columns, operating costs follow the below relation.

79. Conclusions
Capital costs for API separators are heavily dependent on inlet/outlet concentration ratio. Operating costs for API separators are independent of concentration.
Capital costs for activated carbon adsorption are independent of concentration. Operating costs for activated carbon adsorption are heavily dependent on concentration.

80. Conclusions
Both capital costs and operating costs for reverse osmosis are heavily dependent on inlet/outlet concentration ratio.
Capital costs for Chevron Wastewater Treatment are dependent on concentration. Operating costs for Chevron Wastewater Treatment are nearly independent of concentration.

81. Water Management
Questions?