Read Whessoe_tech.pdf text version

Technology & Services

Advanced Liquified Natural Gas Storage Tank Management

a report by

D o m i n i k U z n a n s k i 1 and P i e t e r V e r s l u i j s 2

1. Project Manager and LNG Expert, Cryogenic Studies Section, Gaz de France Research & Development Division; 2. Vice President and Director Sales & Marketing, Whessoe, S.A.


With an increasing diversification of LNG supply sources caused by a growing number of liquefaction plants around the world, an increase in short-term trade and a general trend toward a worldwide liberalisation of gas markets, reception terminals need to deal with a greater variety of incoming liquified natural gas (LNG) qualities. With the need to reduce capital and operating costs, existing storage capacity must be used to its maximum extent and the capacity of new storages must be optimised. It is within this context, and as a response to the previously mentioned trends, that Gaz de France has developed a software model called LNG MASTER®, which can accurately predict the behaviour of LNG in storage tanks. As such, LNG MASTER has three main areas of application: · it helps the operator to analyse and select the optimum tank loading procedure for a new, yet to arrive LNG cargo; · once stored, the software can be used to evaluate the evolution of the LNG in a terminal's tanks in view of optimising their handling, in particular for stratification evolution and LNG send-out operations, among others; and · it can be applied by tank designers to optimise the design of new sites. Gaz de France itself found its use invaluable when it agreed to receive LNG on a regular basis, from 1999 onwards, from Nigeria for the ENEL Italian company via the Montoir-de-Bretagne LNG receiving terminal. This article will examine the French context with a diversification of LNG sources; the challenges that are often faced by the terminal operators who need to handle different qualities of LNG concurrently in


the same storage tanks (effects such as stratification, rollover, or flashing will be explained and some practical rules to manage such cases will be given); and the LNG MASTER software will be discussed, as developed by Gaz de France ­ a real operational tool allowing operators to optimise LNG tank management. In conclusion, typical case studies will be presented as examples of the use of LNG MASTER at the design stage or when operating an LNG terminal.

French Context and Diversification of LNG Sources

The level of imports are increasing at the two French LNG receiving terminals within a context of diversification of LNG sources, in particular at the Montoir-de-Bretagne terminal where LNG tank management has been optimised while simultaneous modification works at the terminal have been undertaken. A long-term contract was signed in 1992 between the Nigeria LNG Company and Gaz de France for the reception of 0.5bcm/y of natural gas. Moreover, Gaz de France signed a `swap' contract with ENEL in 1997 to receive Nigerian LNG in exchange for gas redeliveries from Gaz de France's existing basket of supplies. Figure 1 shows these swaps. As a consequence, the Montoir-de-Bretagne LNG receiving terminal, on the Atlantic coast of France, required some adaptations in order to receive these additional quantities of LNG.

Challenges Faced by the Operators

When mixing different LNG qualities in the same tank, terminal operators have to face three cases. · The first case is `safe' to operators. The incoming LNG is lighter than the LNG in the tank to be filled. A tank bottom filling operation ensures a complete mixing of the two LNG qualities (the LNG injected at the bottom being lighter than the stored LNG) and there is no risk of creating a


Technology & Services

Figure 1: ENEL `Swap' Contract

transferred from the LNG carrier to the filled tank is limited by the hydrostatic pressure at the bottom (due to the liquid column of stored LNG). · In the second case, the incoming LNG is heavier than the stored LNG. A tank top filling operation avoids stratification and the risk of subsequent rollover but usually results in excessive vapour evolution, due to the flashing of the injected LNG in the tank's vapour space. · The last case is the situation that is the most difficult to deal with and needs care. Tank bottom filling with an injected LNG heavier than the stored LNG can lead to a stratification, which will then need to be managed, in order to avoid the rollover risk. This stratification creation risk depends on the injection speed, the density difference between the injected and stored LNGs, the geometry of the filling device and the height of stored LNG. Figure 2 illustrates the second previously explained case, showing the BOG production rate and operating pressure increase during a tank top filling operation. In the example shown in Figure 2, heavy LNG is top filled over light LNG in one of the tanks at the Montoir-de-Bretagne receiving terminal in France. The measurements show the time evolution of the total gas flow rate exiting the tank and of the gas phase pressure. As can be seen, a significant amount of gas is flashed off during the filling operation. The operating pressure of the tank rises steadily during tank filling and then decreases after the end of filling. The operator has several options in order to mitigate the consequences of tank filling and reduce their effects. Firstly, the tank filling flow rate can be reduced in order to decrease the liquid level displacement effect (piston effect) and the gas phase flashing rate. However, this might not be possible due to constraints relative to the maximum staying time of the carrier at the berth. A preferable solution is to regulate the tank's operating pressure in a judicious way, in order to minimise gas production during tank filling. This can be achieved by initially pre-cooling the tank heel before unloading occurs by lowering the operating pressure (this draws-off more BOG, thus lowering the LNG temperature). Just before unloading, the operating pressure is increased above the nominal operating pressure, in order to limit the amount of flashing of incoming LNG coming into the tank's gas phase. This new operating pressure is then maintained throughout the filling process. Once tank filling is achieved, the pressure is then lowered progressively to the initial nominal value.


Figure 2: Total Gas Flow rate and Pressure Evolutions During Tank Top Filling at the Montoir Terminal

Figure 3: Pressure Optimisation of BOG Rate During Tank Top Filling


stratification that can potentially lead to a rollover. The boil-off gas (BOG) production, which can be generated due to the temperature rise of the LNG

Advanced LNG Storage Tank Management

Figure 3 shows a comparison between the total BOG rate generated (surface evaporation and gas phase flashing of incoming LNG) with and without operating pressure optimisation when heavy LNG is top filled on top of light-stored LNG at a filling rate of 10,000m3/h. The results were obtained using an LNG MASTER computer simulation. The results show that by modulating the operating pressure, the total BOG rate generated during filling can be reduced by approximately 50% in this example. This highlights the advantages that can be reaped from this procedure, not only in terms of cost savings by reducing compressor and gas heater output, but also in terms of safety by avoiding the use of safety equipment, such as site flares. In the third previously mentioned situation, the creation of a stratification consecutive to bottom filling of heavy LNG under light-stored LNG and its management after tank filling has ended, can lead to significant reductions in BOG generation, when compared with top filling operations. This is illustrated in Figure 4, which shows the gas flow rate treated by the terminal's BOG compressors for tank top and bottom filling of heavy LNG into light LNG, with and without operating pressure optimisation. However, once such a stratification is formed, it needs to be managed safely. A good prediction of stratification evolution is needed to carry out such management. Figure 5 shows a stable stratification in a storage tank. The heat inputs into the bottom layer gradually increases its temperature and reduces its density. If the system is left to evolve unmitigated, the density difference between the two layers vanishes and leads to a rapid mixing of the two layers ­ this phenomenon is called rollover. Due to the fact that the bottom layer is superheated with respect to conditions in the vapour space at rollover, this phenomenon is accompanied by a transient high rate of vapour evolution that can be 10 to 30 times greater than the tank's normal boil-off rate, thus giving rise to a hazard, due to the potentially harmful overpressures the tank can experience. The first signal of a stratification's presence in a tank is a decrease of the boil-off rate of the tank and an increase of the temperature of the LNG in the bottom part of the tank, because the heat leaks in the bottom layer are not evacuated at the free surface by evaporation but contribute to that layer's temperature increase. Figure 6 (a) and (b) show the evolution of an LNG stratification created in a 500m3 LNG tank during


Figure 4: Comparison of BOG Generation for Top and Bottom Tank Filling Operations

Figure 5: Stable LNG Stratification Behaviour

the Gaz de France experimental campaign conducted between 1987 and 1989. This evolution can be broken down into four distinct phases, each governed by its own phenomenology. During a first phase, the stratification's layers can be considered as insulated both heat and mass wise and only the lower layer heats up progressively, which decreases the density difference between the layers. During a second phase, interlayer penetration takes place between the two layers, reducing the layer's density difference even more. During the third phase, density equalisation occurs, which results in a rapid mixing of the two layers, producing the rollover event. The rollover is characterised by a sudden liberation of overheat accumulated in the lower layer, which can now be liberated at the free surface through evaporation. The LNG then progressively loses this overheat and returns to an equilibrium state in a fourth phase.


Technology & Services

Figure 6: Stratification Temperature and BOG Rate (a) and Density and BOG Rate (b) Evolutions

Boiloff rate (n3(n)/h)

Bioloff rate (n3(n)/h)


Figure 7: Association of LNG Behaviour Predictive Software and Tank Inventory Software

· the prediction of LNG aging; · the prediction of LNG tank filling operations; · the prediction of stratification evolutions and the rollover phenomenon; · the coupling of boil-off rate with the operating pressure and its variations; · the taking into account of tank and site gas recovery and safety devices (compressors, flare, safety valves, rupture disk); and · the specification of operating scenarios for the pressure, the unloading of methane carriers, the site emission and LNG transfer between tanks. As previously mentioned, the optimum cargo discharging/tank loading scenario is selected from various simulations run on the software prior to the actual arrival of the carrier. Once the tanks are loaded, the actual storage must be monitored and compared with the simulations previously made. If so required, new predictive simulations can then be run, using the actual data obtained through the tanks' instrumentation system. The system applied by Gaz de France is Whessoe's Total LNG Storage Instrumentation System®. Currently, Whessoe's SCADA platform (LNG MANAGER®) and LNG MASTER still run independently; however, studies are underway to eventually integrate both, in one single software. The association of the Model 01146 travelling liquidtemperature-density (LTD) gauge from Whessoe with the LNG MASTER software gives the operator an integrated predictive tool with realtime validation, in order to optimise the management of LNG storages. Whessoe's Total LNG Storage Instrumentation System collects all the data from the storage tank and displays this on their LNG MANAGER SCADA platform. Gaz de France operates Whessoe systems at the Fos and Montoir LNG terminals and in the 500m3 LNG tank at Nantes, when it was still in operation. Competing systems from other companies for LTD measurements in LNG tanks, and for predicting tank stratification/boil-off behaviour exist. However, Whessoe's Total LNG Storage


Figure 7 also shows the difference in nominal BOG rates between the stratified tank and the tank after rollover and return to equilibrium has occurred. In the trial shown, the stratification reduced the nominal BOG by a factor of five. This and other examples show the advantages in using stratifications to decrease electrical power consumption for BOG gas compression during LNG aging, and thus terminal operating costs; however, LNG stratifications need to be managed in total safety. The technique of using state-of-theart LNG handling software tools in order to determine a stratification's critical emptying rate above which rollover can be avoided, is illustrated in this article.

Presentation of LNG MASTER

LNG MASTER is a Windows-based, commercially available software with an on-line `help' facility that has been designed to be `user-friendly' for operational engineers. LNG MASTER helps to optimise the handling of LNG storages in terms of safety and cost reduction. Its main functionalities are:


Density (kg/m3)



Advanced LNG Storage Tank Management

Instrumentation System, when combined with Gaz de France's LNG MASTER software, is the only truly integrated system that allows realtime comparison between measured and calculated LTD profiles. The measured LTD profiles after tank filling can be fed into the model and the predictions validated and adjusted accordingly. In this way, LNG MANAGER allows regular updating of tank contents and LNG MASTER carries out the predictions of their evolutions, thus enhancing safety and operating flexibility. A three day on-site training package is performed by either Whessoe or Gaz de France for those companies who purchase LNG MASTER. The use of LNG MASTER can help with the safety aspects of tank management. With a good model for predicting stratification behaviour and rollover occurrence, it is easier to design the emergency relief valve capacity and to persuade safety authorities that it is adequate. The accurate prediction of the behaviour of LNG in storage tanks using LNG MASTER also helps to reduce capital expenditures. The capacity and the number of new storages can be optimised; for example, in case of extension of a LNG receiving terminal, existing storage capacity can be used to their maximum extent, and the capacity of new storages can be minimised, thus reducing cost. The boil-off recovery system (number of compressors, capacity of re-liquefaction unit, etc.) can be optimised by making a good prediction of the boil-off rate. With an accurate prediction of stratification and rollover, it is also easier to design the safety devices. The operating expenditures relative to boil-off handling can be reduced using LNG MASTER: · by a judicious use of tank operating pressure, boiloff rate can be significantly reduced; and · by a deliberate creation of stratifications, high boil off rates can be avoided during tank filling. An appropriate model such as LNG MASTER is absolutely necessary in order to predict the rollover risk and how to avoid it, even if this method induces some constraints on daily terminal management (in term of send-out flow rate, carrier supply frequency and minimal storage levels). Another application of LNG MASTER lies in optimising compressor output as a function of electricity prices (daytime versus night-time). By raising the operating pressure during daytime and lowering it at night-time in a periodic way, savings in electrical costs can be made.


Figure 8: Daily Electricity Cost Savings with Optimised Operating Pressure Management

Applications of LNG MASTER

Modification of an LNG Terminal for the Import of a New LNG Quality

The modifications of the Montoir-de-Bretagne LNG terminal for the reception of LNG from Nigeria are presented as an example of the use of LNG MASTER at the design stage. The Montoir-de-Bretagne LNG terminal was built in 1980 and has received LNG from Algeria since the beginning of its operation. With the ENEL contract, the Montoir-de-Bretagne LNG terminal has to handle three different LNG qualities (two Algerian and one Nigerian grade). Simulations made with LNG MASTER showed clearly that using existing storage tanks without dedicating a new one was absolutely possible, provided that the BOG recovery systems were reinforced. Due to composition differences, the mixing of Nigerian and Algerian LNG in a storage tank may create additional BOG during unloading, particularly when filling from the top is required to avoid stratification. The BOG is compressed and incorporated into send-out LNG in a re-condenser, as commonly performed, but the capacity of the recondenser was not sufficient in the new situation to avoid the flaring of gas. The feasibility of mixing Nigerian and Algerian LNGs as well as the resulting new boil-off rates were estimated using the LNG MASTER software. Regarding these results, the following items were modified to reduce environmental impact:


Technology & Services

· a new re-condenser was built with a larger capacity; and · three cryogenic were added. boil-off gas compressors

Figure 9: Stratification Critical Point

Sizing of BOG Compressors with Respect to a Worst-case Scenario

The BOG compressors for a new terminal to be built have been sized by considering a worst-case scenario consisting of a typhoon inducing an atmospheric pressure decrease of 15mbar/h over six hours. The effect of such a depressurisation was calculated with LNG MASTER. The results obtained gave valuable information on the transient generation of BOG and allowed realistic sizing of the compressor units to be installed.

Adaptation of Boil-off to Electricity Pricing for Homogeneous LNG Aging

Figure 10: Stratification Critical Emptying Rate

The operating pressure is a key factor for transient control of boil-off in LNG storage tanks. By varying the pressure, it is possible to anticipate on the tank behaviour, thereby optimising the daily management of boil-off phenomena. Optimisation can involve savings in capital expenditure and operating costs related to the compression function. Figure 8 shows a generic example of optimisation, concerning the adaptation of boil-off flow rate to electricity prices. Electricity contracts generally provide pricing according to specific time slots reflecting the conditions of power supply and demand. Depending on the contracts, prices may vary seasonally or daily. By varying the operating pressure, it is possible to shift the power consumption of the compressors during low-rate time slots. For example, in the case of daily cycles, the pressure is allowed to rise during the high-rate period, which reduces the boil-off and hence the related power consumption. Conversely, during low-rate time slots, a drop in pressure will release the previously accumulated overheat by generating additional evaporation.

Optimisation of LNG Stratification Management

tank filling during LNG aging can also be significantly reduced. However, in managing LNG stratifications, the operator needs to be aware of the basic behaviour of stratifications and also be able to anticipate their evolution up to the rollover, in order to take timely appropriate remedial actions to avoid this phenomenon. Among the rollover mitigation methods available to the operator, stratified tank emptying represents one of the surest and most economical methods to safely manage stratifications. It is the method of choice for LNG terminals, since they have to emit their LNG continuously and can therefore distribute the terminal's emission requirements on the different storage tanks in an optimal way. The emptying flow rate necessary to avoid rollover occurrence must be sufficient to completely empty the lower layer of a stratification before its density equalises with that of the upper layer. The effect of lower layer emptying on the evolution of a stratification translates into a reduction of the rollover onset time. As the emptying rate increases, the rollover occurs earlier, since the tank wall heat leak into the lower layer is attributed to a smaller and smaller volume. However, above a given emptying rate, the lower layer is entirely emptied before rollover occurrence.



The intentional creation of LNG stratifications can be an interesting alternative to the complete mixing of cargo and tank heel LNGs. By deliberately creating a stratification, in particular in the case of filling of heavy cargo under light heel using a tank bottom filling device, the operator not only avoids high BOG production rates during the filling operation, but the nominal BOG rate after

Advanced LNG Storage Tank Management

Figure 9 and 10 illustrate the notion of critical emptying rate and critical point of a stratification and present an application of this concept to an industrial stratification using Gaz de France's LNG MASTER software. Figure 9 shows the emptying rate curve giving the time necessary to empty the lower layer of a stratification at the prescribed emptying rate. The rollover time curve represents the rollover onset time at the given emptying rate. As the emptying rate increases, the rollover onset time decreases. At sufficiently high emptying rates, the two curves intersect. This intersection, defined by the critical onset time and the critical emptying flow rate, defines the critical point of a stratification. The critical emptying flow rate of a stratification is the flow rate at which the lower layer is entirely emptied just as rollover occurs. Operating at an emptying flow rate above this critical flow rate ensures the withdrawal of the lower layer before rollover occurrence. In this way, the region in Figure 9 to the right of the emptying rate curve represents the safe operating zone of a stratification. The region to the left of this curve represents the danger zone of a stratification. The critical emptying flow rate constitutes an interesting criterion to evaluate the compatibility of a stratification with the site's operational constraints, relative to the emptying capacities (number of submerged pumps) of the stratified tank. If, for a given stratification, the operating conditions and tank characteristics allow the use of an emptying rate above the critical rate until the lower layer has disappeared, then the stratification can be managed in complete safety by the site's operators. Figure 10 shows this concept being applied to a 10m GL1K upper layer/25m GL1Z lower layer stratification in a 120,000m3 (63m diameter) tank. The rollover time curve in Figure 10 has been constructed with a series of seven LNG MASTER simulations. The results show that this particular stratification can be managed in total safety by using two submerged pumps of 450m3/h nominal output, which is entirely compatible with large scale tanks that are usually equipped with three or four submerged pumps. This type of parametric study using systematic simulations can be repeated for various stratification configurations. Figure 11 illustrates the results obtained by varying the stratification layer heights ratio, while keeping the total LNG height constant at 35m. The results highlight some interesting tendencies. Firstly, all stratifications studied can be managed safely in a tank equipped with three


Figure 11: Various Stratification Critical Emptying Rates

submerged pumps of 450m3 capacity. Secondly, the rollover time evolution curve shows that the longest stratification management period (approximately six days) is obtained with an equal layer 17.5m/17.5m stratification, which can be managed with just one 450m3 pump. The examples previously presented highlight a technique that can be used to optimise the safe management of LNG stratifications. In order to apply this technique safely, accurately advanced LNG stratification prediction tools such as LNG MASTER, which the operations engineer of shift operator can use easily are indispensable for ensuring the smooth operation of the site.


In the current context of diversification of LNG supply sources and the liberalisation of gas markets, operators of LNG installations as well as engineering firms need tools to allow the optimisation of LNG storage tank management and the design of LNG installations. The LNG MASTER software developed by Gaz de France can help achieve such optimisation, in particular for minimising BOG generation in daily terminal operations and for sizing the BOG recovery and safety devices of the terminal and its tanks. The examples shown in this article illustrate some of the techniques that can be implemented with the help of LNG MASTER and highlights the advantages of safely managing LNG stratifications for reducing operational costs. We are confident that LNG MASTER can successfully help gas companies, operating LNG storage sites and engineering firms face the challenges of today's rapidly evolving LNG industry more effectively.



7 pages

Find more like this

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate


You might also be interested in