Storage tank loading schedules at LNG terminals. A dynamic simulation approach
Rising global gas demand driven mainly by cleaner power production and the availability of new natural gas sources are driving significant growth in the LNG markets. Large natural gas field discoveries on and offshore have prompted several countries to plan liquefied natural gas (LNG) export projects, including in North America, Australia, East Africa and the east Mediterranean. In one of such scenarios the expansion called for two new 50,000 m3 paraffinic naphtha storage tanks to increase the capacity and to replace existing tanks, which had reached the end of their design life.
The project was commissioned to process additional quantities of associated low-pressure gases, which became available with the oil production increase.
Due to the light components content of the product, the storage and shipping of the naphtha require the temperature of the product not to exceed 20 °C . A dynamic simulation study was performed to evaluate conditions during the filling, storage, and unloading stages with enhanced representations of heat transfer within the storage tanks.
Aspen HYSYS version 7.3 was used to model the process in dynamics mode with additional heat transfer correlations coded and linked to the model to evaluate the heat transfer impacts on the product temperature at one-hour intervals.
Figure 1. Configuration of the dynamic simulation model in Aspen HYSYS
With the given product composition and equipment dimensions (including the insulated sections of the pipeline) the model was also adjusted for diurnal ambient temperatures as shown in Figure 2.
Figure 2- Ambient temperature
Loading and Unloading Schedules
A number of scheduling options were reviewed to evaluate the impact on temperature of changing flow rates feeding both tanks either in series of simultaneously. The schedules included standby periods before unloading the product to the ships.
To model the temperature rise in the worst-case scenario, it was assumed unloading of the tanks would start at 10AM. Unloading of the tanks would initially start at a lower rate to limit the flow velocity to 1 m/s and avoid accumulation of static electricity charge in the tank. After 7 hours, the discharge rate is ramped up until fluid level in the first tank reaches its Low Low Liquid Level (LLLL). After that, the second tank is drained at the same rate of 2000 m3/h down to its LLLL.
Heat Transfer Models
The heat transfer mechanisms involved in the system were modelled and calculated in a separate programming routine and the resulting values exported to HYSYS at each time step. These calculations include heat transfer due to radiation, convection and conduction. Figure 3 shows the details of the heat balance in each wall of the tanks. Note that there is a secondary containment wall that creates an air gap between the environment and the tank wall. The heat transfer through the tank floor was negligible compared to the other modes of heat transfer and, therefore, was not considered.
In initial studies, HYSYS pipe heat transfer was used to model heat transfer of the fluid in pipes. Later, it was requested to develop a more comprehensive model to include effect of solar radiation in pipes. Figure 4 shows the heat balance used to calculate heat transfer in pipes.
Results and Conclusions
The results of the combined HYSYS- heat transfer dynamics model enabled the estimation of the temperature profiles for each one of the tanks and throughout the pipeline system.
Figure 5 illustrates the results obtained for a particular schedule in both tanks. It can be seen that following this schedule the product remains under the required temperature constraint during both the loading and unloading events. The product temperature rises very close to the limit when a tank is left in standby at LLLL while the other tank is being loaded. Higher temperatures may be actually seen if longer standby periods are experienced. However, the model predicted that if both tanks were loaded simultaneously, the liquid temperature will remain at safer levels.
Not shown in the graph but also of interest is that the temperature of product received at the ship would initially exceed the temperature limit regardless of the time of unloading (day or night) due to the presence of an uninsulated pipe segment. Therefore, it was recommended that the entire piping between the tanks and the jetties be insulated properly.
This case study shows how a dynamic simulation can be enhanced to include other heat transfer phenomena and deliver important insights for the design and operation of these facilities.