RFCCU Technology - Power Recovery Turbine (PRT) Design Considerations

In the initial conception of a RFCCU project there are two foremost considerations: The design feedstock and accompanying yields, that drive the products margin, and whether to use a Power Recovery Turbine (PRT) to recover power from the regenerator fluegas or stay with only heat recovery for steam generation. Thus, one can consider these as boundary conditions, since all other equipment in-between is standard for all technology providers, perhaps with some minor differences, and the sensitivity to product margins, in my opinion, is minor

This article is only for educational sharing of my opinion and can have inaccuracies and omissions. This is not professional advice for any real situation of any kind. Readers assume all risks for any deduction or extrapolation. The Disclaimer at the end of this Article should be reviewed. It applies to the entire content and any comments or response

For the sake of this Article, let us assume that the feedstock and the design case with max CCR and metals limits has already been decided, and the project team is now faced with the task to determine Yes/No for the power recovery option. This is a critical decision prior to prequalify the technology providers and their alliance vendors for the PRT, prepare the RFP and the design considerations and what information must be submitted by the bidders for owner's evaluation of options, technology and reliability

So where to start? The first thing to consider is how the utilities, mainly the HP steam & Power, is planned to be supplied and the anticipated price? There are normally two main options: Purchase power over the fence (OTF) and produce needed steam on-plot; Or produce both power and steam on-plot using combustion gas turbines in a combined cycle mode. Next, comes the estimation of the total power requirement of the unit. This is guesswork at this juncture in the project, but if the project process engineers have some done preliminary planning work to estimate the overhead products flow from the main fractionator and the required air to burn off the expected carbon (to delta coke level) and sulfur on catalyst, friendly vendors of Axial Air Blowers and Wet Gas Compressor can provide an estimated number for the rated HP/KW. This leaves how much the rest of the unit may need? In my opinion, a starting guess can be approximately same as the total of the what is needed for the for the two big rotating machines

The next step is to estimate how much KW can a PRT may produce basis the design fluegas rate and its temperature and pressure, and the approximate installtion cost of the PRT section equipment & piping. In my opinion, it is best to inquire from the vendors for both the single and two stage machines so the differences can be reliably quantified for decision making. However, if vendors are not cooperating, then either rules of thumb or public domain information are other recourses. One, that I recall for power, was around 1 MW for each 4000 BPD feed, meaning, e.g, for a 60,000 BPD unit approximately 15-20 KW range

As for the cost it is more difficult to estimate because a PRT installation requires several things besides the machine itself: A third stage separator (TSS) to remove catalyst particles to the desired size cut from the flue gas; A fourth stage separator to handle underflow from the third stage, piping, electrical and controls hardware to and from the machine, use of alloy material, for hot wall design, in order to prevent any refractory particle go in the machines and a bypass and safety systems to operate the when the PRT is down. So, this number need some intelligent estimation. But as a guess, in my opinion, if in rush, around 10% of the cost of the unit may be a conservative initial starting point until better numbers are available from internal or external sources. For meticulous work, it may be worth paying a contractor to estimate this cost  

However, before a model can be run, assumptions on the reliability of the machine are also important. PRTs’ are around for about 40 years and have evolved. During the early years the reliability, I believe, was inadequate to align with, say, a 4-year T/A. However, it has considerably improved over years, so, in my opinion, it is best to talk to owners’ who have recently installed, say, in the last 5 years, and get feedback. Though vendors can also provide the calculated MTBF, but I think it is built from components MTBF. So, the actual data, I think, is more reliable because of exposure to fluctuations in the flue gas rate, P&T, and the dust loading (which, I think, can sometimes be a cause of excessive vibration, due to deposition of catalyst on blades, and may need S/D to clean the deposits). The reliability data is important because net margins from the RFCCU are significant and thus cash flows need to be adjusted (-ve) in the NPV calculation due to unplanned emergency shutdowns.

Other adjustments that one may consider are: The additional cost of specialty catalysts, like ZSM or DeSox or Propylene selective or matrix cracking, lower coke production, etc; The need for a catalyst cooler, if technology provider requires it, for the estimated coke make including both the CCR of the feed & the catalytic coke production; Higher O&M costs and an estimate of power and steam cost adjustments in the model, when PRT is running; Other adjustments may be credits for lower carbon emissions, if applicable, energy efficiency, such as the Solomon's Index and other potential credits that mat be locally available.

Also, the amount of sulfur in coke in the RFCCU units present another challenge. If only DeSox catalyst is used, its effectiveness is limited, my guess is around 60-70%, which means that flue gas may contains SO2 and to a lesser extent SO3. These compounds, I think, can attack the Nickel containing alloys in the machine, if applicable, and can lead to sulfide stress cracking and cause corrosion, due to formation of Nickel Sulfide, in the long run. And in cases where water or steam quench may be used upstream of PRT to control temperature this affect, in my opinion, may get even more pronounced through production of Carbonic and Sulfurous acids 

Thus, there are many considerations before a decision can be made on PRT. The above are just a few selected ones for the purpose of initial screen. But, refiners' project development staff, process engineers and relationship vendors must work together to look at all factors, qualitative and quantitative, to arrive at a well considered and rational decision, The bottomline purpose should be to augment the products margin in an economical way - lowest cost feed to make highest margin products with minimum capital and operating cost. Local factors, such as power availability, staff skillset, vendor support arrangement in the countryon and reliability considerations. This article is meant for a new grass-roots RFCCU, but if PRT needs to be added to an existing brownfield refinery, then a lot more work, I think, needs to be done with respect to the utilities interfaces and the impacts on the remainder of the refinery

Disclaimer: This is not professional advice, directly or indirectly, and can be edited or deleted at any time. Anyone accessing this Blog unconditionally agrees that the expressed views are only the personal opinions of the Author for educational sharing only, basis author’s knowledge only, and may contain omissions and inaccuracies. It must not be used for any actual new project or on an existing facility. Readers accept all risks and responsibility for any interpretation or extrapolation and any consequences stemming from such reliance