PTQ Q3 2023 Issue

215 19

250 psig steam

Vacuum tower (process load)

125 psig steam

1st stage ejector

2nd stage ejector

3rd stage ejector

New equipment Pressure mmHg absolute Temperature, ˚F

CW supply

1st stage intercondenser

2nd stage inter- condenser

4th stage ejector

CW return

Non-condensable gas

Hot well

Oil

Steam condensate

Condensable oil (slop oil)

Motive & process steam

Sour water

Figure 6 Vacuum system modifications

operating with a discharge pressure well above the maxi- mum design pressure (MDP). This is commonly referred to as broken operation. Performance breaks are caused when an ejector’s discharge pressure exceeds its MDP, caus- ing it to operate erratically, at high suction pressure, and with very high noise levels due to surging. The first-stage condenser pressure drop was three times design, and the vapour outlet temperature was much higher than design. It could not reduce the second-stage ejector load low enough to maintain the first-stage discharge pressure below its MDP. The inter-condenser outlet temperature sets the second-stage ejector process load; therefore it determines second-stage suction pressure based on its performance curve. The root cause of high first-stage suction pressure was poor performance of the first-stage intercondensers. The original first-stage inter-condensers were fixed tube sheet design with triangular pitch tube layout. Triangular pitch layout is normally avoided because of inherent diffi - culty cleaning the shell-side. For this reason, a square pitch design is usually favoured for this service. However, using triangular pitch enabled the original design to be built using two intercondensers instead of three. More tubes fit in the same shell diameter compared to square pitch design. Significant cost savings in piping and equipment justified the risk of being unable to clean at that time. Based on trou- bleshooting efforts, it was clear that the inability to prop- erly clean the bundle was only partially to blame for the high-pressure operation. The vacuum system was originally designed with 57,000 lb/hr process steam load. This is considered very high. The

combination of high steam load and the required hydrocar- bon load resulted in a design motive steam rate of 122,000 lb/hr. As a result, the first-stage ejectors and condensers were very large. New intercondensers needed to be square pitch design so that they could be properly cleaned. Less surface area could be accommodated in new exchangers determined by length, diameter, and plot space limitations. This meant that the condensing duty needed to be dramat- ically reduced. After thoroughly evaluating the ejector system operation, including the condensers, it was determined that all four ejector stages and new first- and second-stage inter-con - densers were needed to achieve the vacuum column VR yield reduction goals. The original third-stage inter-con- denser and fourth-stage after-condenser could be reused. (see Figure 6 ). A major challenge was to find a way to reduce con - densing load while at the same time providing adequate MDP margin. Motive steam rate is a consequence of the design suction pressure, process load, and MDP. An under- appreciated vacuum system design specification is the MDP margin. The MDP margin is the discharge pressure above the theoretical minimum based on the second-stage ejec- tor’s suction pressure and system pressure drop between the first- and second-stage ejectors. In reality, there is a minimum condensing pressure based on cooling water supply and return temperatures. MDP margin allows the first-stage inter-condenser to operate at higher condensing pressure than design if the exchanger heat transfer coeffi - cient decreases due to the many unknowns, such as cooling

PTQ Q3 2023