Below, I list sources you can turn to when estimating the cracked gas generated from vacuum tower heaters. That way, you know the flowrate and composition of light-ends generated in the vacuum tower’s pre-heat furnace.
Background: Vacuum Units are typically used in refineries to treat heavy Atmospheric Tower Bottoms. (The heavy liquid bottoms leaving the first major distillation column in the refinery, the Crude Unit). In order to distill these very heavy liquids, the vacuum tower is operated at very low pressures: near vacuum levels, hence the name. The low pressure makes it easier for the heavy liquid to flash, allowing distillation.
In addition to the very low pressure, a high temperature is required in the Vacuum Tower. Typically this is supplied first by pre-heat heat exchangers, and then finally by a furnace. The hotter the furnace outlet temperature, shown above as temperature “X,” the easier the distillation will be. However, the intense heat can cause chemical reactions, like coking and cracking. Coking forms hard solid deposits in the heater while cracking breaks molecules into gases. Both cause a loss of liquid product, and can complicate downstream operation. In the past, furnaces were usually kept to ~725-735°F to minimize these problems. In modern times, with sophisticated heater designs to minimize problems, some daring revamps try 785-810°F. 750°F is probably a good “average” temperature to expect these days.
Any temperature in these ranges will still cause cracking. But how much cracking? What composition of cracked gas will form?
The truth is, the only reliable source of information is testing the actual crude that will be run in the refinery. See if someone else is running the crude in their own vacuum unit and can give you data. If not, try a lab test. Or get data from a similar type of crude.
But in the very early stages of a project, you may need to estimate something for the initial design and cost estimates. Where can you turn? A few ideas:
- Operating experience really is best, as noted above
- Process Design for Reliable Operations by Norman Lieberman has a correlation
- SimSci’s Pro-II workbook on hydrocarbon distillation with some very rough ideas
- Refinery Process Modelling by Gerald Kaes has some correlations (section 9.2.2)
- Vacuum tower distillation vendors and vacuum heater vendors may have some data they could share. They could be your third-best source of information, behind operators and lab data
I will mention the simplest idea from the above. It is from Sim-Sci workbook: for a very simple first pass, you can model the cracked gas as 75% ethane / 25% propane in a process simulation and assume the molecular weight will be in the range of 30-35. This will let you design the pipes and other equipment handling the cracked gas. Realistically, other components will be there in the gas; potentially things like hydrogen, methane up to about pentanes, CO, CO2, N2, O2, H2S, etc.