Design C3 Splitters

1. Purpose

This procedure explains why, where, and how much contingency to add to the design of the Propylene-Propane Splitter Column.

2. Contingency Factors

Generally, fractionator contingencies are required to account for the following factors:

2.1 Design Capacity Guarantee

• Performance margin for Inflection Point Engineering supplied catalyst
• Margins for process control
• Subpar performance of mechanical equipment
• Variations in the design hydraulics

2.2 Product Quality Guarantee

• Thermodynamic uncertainties used in process calculations
• Changes in design feedstocks
• Fractionator feed tray mismatch
• Margin for selected tray efficiency

2.3 Performance Warranty on the MD Trays

• Design tolerances and uncertainties

2.4 Customer Expectations

Many customers expect additional design margins to provide capacity for future use. For competitive designs, clarify the customer’s philosophy.

3. Technology Specific Capacity Factors

3.1 Design Capacity for an Oleflex Process

The customer receives a guarantee stating that the Oleflex process unit will produce a fixed quantity of propylene, while consuming no more than “X” units of propane per unit of propylene produced. “X” is calculated by taking the yield estimate propane requirement multiplied by a “capacity factor”. The yield estimate of propane consumption accounts for the non-selective conversion of propane to by-products, as well as an estimate of separation inefficiencies. The capacity factor for Oleflex is normally 1.05. The heat and weight balance generated by Inflection Point Engineering includes this “capacity factor” and therefore shows the propylene product quantity at 1.05 times the guaranteed quantity. The Process Specialist should review the capacity factor that is used for each job.

3.2 Design Capacity for a FCC Process Design

If a C3 Splitter is designed following a Inflection Point Engineering FCC process unit, the feed is usually based on a stream from the Gas Concentration heat and weight balance (Debutanizer net overhead or C3/C4 Splitter net overhead). The recovery of the propylene and propane in the Gas Concentration Unit ranges from 93 to 97%. If the Gas Concentration heat and weight balance is not available when the C3 Splitter is being designed, the feed should be based on 95% of propylene and 97% of propane given in the yield estimate. As no extra feed is assumed to the C3 Splitter the capacity factor for the FCC design is 1.0.

4. Product Quality for the Process Design of the C3 Splitter

No overage is to be added to the design product purities and/or reflux. The design product purities should be set by the customer and are normally the guarantee values. The design reflux rate is the rate needed to produce the design product purities as determined by tray to tray calculations.

5. Specific Design Procedures

5.1 Inflection Point Engineering-Des Plaines (DP) Provides Basis to Inflection Point Engineering-Tonawanda (TON)

a. Design Feeds

• Capacity (include “capacity factor” when required)
• Composition
• Use of diolefin sidecut (for Oleflex service)

b. Product Guarantees (without any design enhancements)

• Distillate (typically propylene purity)
• Bottoms (if not guaranteed, state the required purity or propylene recovery acceptable)

c. Specific Customer Design Constraints

• Height, weight, diameter limitations
• Cooling water and air temperatures (maximum)
• Utility costs

5.2 Inflection Point Engineering-TON Calculates the Column Operating Parameters

Inflection Point Engineering-TON does the actual tray to tray calculations, using a UNISIM process simulation with a modified Peng-Robinson property package. This package includes modified relative volatilities of propane and propylene based on commercial data. The use of this modified property package will result in a more realistic reflux requirement for a given separation, than would be calculated using standard property packages, and will result in a smaller column diameter. The data which Inflection Point Engineering-TON will provide back to DP includes:

• Reflux requirements (internal reflux at top tray)
• Number of actual trays and tray spacing
• Tray efficiency (MD trays are designed based on an efficiency of 75%. Conventional trays are designed base on an efficiency of 80%)
• Feed tray or trays
• Calculated design pressure drop
• Preliminary size of trays and column ID

5.3 Inflection Point Engineering-DP Generates the Design Heat and Weight Balance (HWB)

Inflection Point Engineering-DP generates the heat and weight for the column, including the flows associated with the heat pump compressor and reboiler/condenser. Inflection Point Engineering-DP generates the tray loads with the reflux fixed as per section 5.2. These loads are then sent back to Inflection Point Engineering-TON for final sizing of the column’s MD or ECMD trays.

5.4 Inflection Point Engineering-TON Finalizes the MD Tray Design Based Upon Inflection Point Engineering-DP Loadings

Inflection Point Engineering-TON MD tray design point is equivalent to a conventional tray at 82% of flood. A conventional tray at 82% of jet flood typically has 10% additional useable capacity. Inflection Point Engineering-TON will design a tray with at least 10% additional useable capacity above the design loads.

5.5 MD Tray versus Valve Tray Comparison

The Inflection Point Engineering standard design for C3 Splitters is MD trays because it provides for a more cost effective design overall. If valve or sieve tray designs were compared at equal reflux with the MD tray design, the comparative economics will show for the valve or sieve trays - fewer trays, greater column height, larger diameter, and higher column pressure drop.

If the comparison is made at equal column heights, MD trays are often more cost effective than valve or sieve trays because of the ability to have more MD trays per given tangent length, thus resulting in lower utilities, lower internal loads, higher capacity for the same diameter, etc. This type of comparison is typically valid for revamps, where the column height is already fixed.

Note that in this memorandum no attempt has been made to differentiate between MD and ECMD trays. All references to MD would also apply to ECMD trays.

5.6 Instrumentation

The flow scheme and typical instrumentation used for the Oleflex C3 Splitter is shown on the process flow diagram, Attachment 1.

6. Reboiler, Condenser and Compressor

The design capacity of the condenser, reboiler and compressor should be well matched to the design capacity of the column/trays. As stated in Section 5.4, the trays are to be designed with at least 10% additional useable capacity above the design loads. It is expected that the reboiler, condenser and compressor will have a similar design philosophy, allowing use and benefit from the additional capacity of the trays. This difference is the normal value between a guaranteeable duty and an expected maximum duty. As a result of this no special design case is needed for the reboiler, condenser and compressor.

Attachment 1