Dover Project consideration about process design
- 1. Heat recovery, Emulsion Trim Cooler and Emulsion cooler outlet temperature control I am not sure this project has proceeded heat integration with Pinch technology or not, but It seems that there is big room for raising the level of heat recovery. I have two questions about current heat exchange flow scheme as below: 1. The outlet temperature of Emulsion/BFW exchanger is set at the required FWKO inlet temperature and no heat duty for trim cooler normally in simulation. The adjustment of inlet temperature of FWKO in operation can be executed in this way: • If the outlet temperature is lower than required operating temperature of FWKO, then part of BFW will need to bypass the exchangers. It will result in lower BFW final temperature and heat recovery level. • If the outlet temperature is higher than required operating temperature of FWKO, then the control valve for cooling glycol to Trim Cooler will need to be opened to cool emulsion down with Trim Cooler. As the Trim Cooler is designed for upset conditions and has quite big design flow (215 m3 /hr), any small glycol flow could easily fall into the dead band of glycol control valve, this make accurate control of FWKO inlet temperature impossible. 2. The outlet temperature of Produced Gas /BFW exchanger is 151°C for Light Gas Condensate case and 157°C for Heavy Gas Condensate case (see attached PDF file) and the latent heat contained in PG could be further recovered by BFW Based on two reasons above, I made a quick study in heat recovery flow and propose some change for current flow scheme as below: Current Flow BFW from PW/BFW Exch PG from Pipeline 2X65.9 m2 H=2.75MW 2X657.1 m2 H=44.8MW 105.8°C 152°C 150.9°C 65°C Glycol 153.6°C 137°C BFW to Steam Generation 9X548.5 m2 H=70.4MW (3P3S) Emulsion from Pipeline 169.6°C 3X105.5 m2 H=0 MW 154.6°C 125.1°C 125.1°C Glycol
- 2. Compared to current flow, proposed flow has some advantages list below: 1. The final BFW outlet temperature is 156.6°C, 3°C higher than current design without any heat transfer increase (6699 m2 for current design vs. 6645.6 m2 for proposal). This means 5.3 MW more heat gain in BFW heat recovery and same amount less glycol cooling duty, which account for 5.9% of BFW heat recovery. 2. This solves FWKO feed temperature unsteady control problem. 3. It eliminates the requirement for two control valves (one flow rate adjustment for BFW through Emulsion/BFW exchangers and one bypass flow control) and maximizes the heat recovery. In current design, 40°C cooling glycol is used as Trim Cooler cooling medium to cool emulsion above 120°C. The potential problem with it is the low wall temperature on the tubes of Trim Cooler and it will cause emulsion density to layer and promote rag layer buildup in FWKO. I am thinking to add a medium level cooling glycol, say glycol at 75°C or 80°C, to be used as emulsion and PW Coolers. Thus, the wall temperature can be raised and rag layer build-up problem can be relieved to some extent. More importantly, some hot cooling glycol in lower temperature can directly be used as cooling medium, rather than being sent to air cooler cooled to 40°C. The advantage for this is that it reduces the cooling duty of air cooler on one hand. On other hand, it raise the hot cooling glycol return temperature and this hot cooling glycol either can be used as heating glycol or be cooled in air cooler with bigger heat transfer temperature difference, so that the air cooler area can be reduced further. Moreover, I found there is no direct temperature control to FWKO feed, the operating temperature of FWKO is adjusted and controlled by TIC-0141 (which measures wet emulsion temperature in FWKO outlet, then adjusting BFW bypass flow or cooling glycol flow). As it is known, FWKO has very big capacity and the retention time of wet emulsion in FWKO is about 30 minutes. Any high temperature PG from Pipeline 3X505.2 m2 H=37.5MW 152°C 120.8°C 65°C 105.8°C 169.6°C 130.2°C 156.6°C 144°C 125.1°C 3X203.7 m2 H=10.1MW 6X623.4 m2 H=40.9MW (3P2S) 3X259.5 m2 H=29.4 MW BFW from PW/BFW Exch Emulsion from Pipeline BFW to Steam Generation
- 3. and low temperature will not be detected by TIT-0141 and TIT-0143 immediately (30 minutes delay for complete detection though they can measure diluted temperature some time later). This control is not sensitive and accurate for operating temperature control in FWKO). As my understanding, though the operating temperature in Treaters is important, it does not mean the temperature in FWKO can be ignored. On the contrary, the gravity separation is more sensitive to temperature change and temperature control is more important in FWKO than in Treaters. On the other hand, if the temperature in FWKO is well controlled, then the temperature in Treaters will be more steadily controlled as the amount of diluent added to FWKO and Treaters are comparatively fixed. This is my preliminary consideration, I am not sure it is reasonable and feasible or not. Which is better, One FWKO or more? One FWKO is used in current design, it may save some investment, but it may not be better considering the turndown ratio, potential Rag layer problem and emulsion flow in pipeline during steaming and early production. As literature shown, typically 3:1 to 4:1 turndown is still the practical limit on heavy oil equipment turndown, cooling and low flow zones will cause emulsion density layering and interface rag layer builder-up, so the bigger FWKO is not always better. Another issue with one bigger FWKO is sand collection in low point of pipeline which will be brought out with low flow during steaming and early production as emulsion flow is required above a critical flow so that sand carried with emulsion could not be settled down. The bigger design flow, the more serious sand settles down during low flow. Heating glycol return from Combustion Air Heater Based on simulation, the heating glycol return temperature is 25°C in winter (the temperature would be around 40°C in summer based on 15°C approach temperature). This way, is it necessary to mix this part heating glycol return with cooling glycol and heating glycol from other sources to pull inlet temperature of air cooler and LMTD down? Why not it is used directly as cooling medium? PSV relieve destination and Flare Header sizing In this project, it seems all PSV outlets are connected to flares. As other project did, relieves from some PSVs were not connected to HP Flare, like water or steam relief, it can be relieved to safety place near equipment and then liquid can be collected in sump and vapour is vent to air without any harm to environment. I am not sure if there is any regulation to rule out these kinds relieves. Another issue regarding flare system is the size of HP Knock-out Drum. The governing case for HP flare header and HP Flare Knockout sizing is wet emulsion relief from FWKO when PW outlet in FWKO is blocked out, thus a 24” header and a huge HP Flare Knockout Drum (15’ID x 80’ S/S, very close to size of FWKO) are designed to handle the relief for the case. A CHD Vessel at same size is designed as spare HP Flare Knock-out Drum. As FWKO has two equal size PSVs, one with lower set pressure designed for fire case relief and outlet block-out case and another with higher set pressure designed for outlet block-out case only. Two PSVs share the total relief load of governing case. In current design, both PSVs are relieved to HP Flare Knockout Drum and HP Flare Knock-out Drum is designed to store relief liquid for governing case. I have been exploring the possibility to connect the PSV for both fire case and liquid relief case to HP Flare Knock-out Drum and connect to another PSV to CHD Vessel, let two vessels share
- 4. the total liquid load to reduce the sizes of both HP Flare Knock-out Drum and CHD Vessel. I am not sure whether this is practical and feasible or not.