A. Taylor, Greenfern Dynamics, Adelaide, Australia; T. WILSON and R. DAVIS, ConocoPhillips Australia, Brisbane, Australia; and F. PIETERSE, Greenfern Dynamics, Brisbane, Australia
The use of advanced process control (APC) technology on LNG liquefaction trains is well accepted with some of the earliest implementations developed in the late 1990s. Several LNG production sites seeking to extend APC benefits have more recently applied the technology to the LNG storage area, often referred to as the boil-off gas (BOG) system. Depending upon the nature of the overall plant design, the BOG from the LNG storage area can be routed to fuel gas, a domestic gas supply or back into the LNG liquefaction process. A spin-off of these design variants is that the optimization objective can be to either maximize BOG yield (to maximize LNG production), target a specific BOG yield to close a mass balance with minimum losses (e.g., flaring) or minimize BOG yield to unload the refrigerant systems driving the liquefaction process.
Although the nature of these BOG APC applications can be relatively simple (i.e., small in dimensions), the interaction with regular discrete ship loading activities is a challenging complication whereby the duty impacts of cooling down loadout lines and ship storage tanks can return significant load disturbances. In effect, there is some element of sequence logic that must be spliced into the continuous control nature of the APC application to achieve the best outcome.
The shiploading events also add challenge to the automation needs in that there can be a relatively high degree of variability in the following:
One of the primary objectives of the BOG APC is to exploit the surge capacity in the allowable LNG tank pressure range to attenuate the effects of ship loading on the process—this is important to get right for those process designs where the mass balance is closed by either the LNG rundown temperature target or BOG flow return to the condensation loop (both impact cryogenic refrigeration needs).
One important aspect of the LNG tank pressure response is that it is not a perfect battery, where in a perfect battery, the net sum of the work saved (i.e., the reduction in BOG vapor at higher pressure) is expended to re-establish lower pressure conditions for ship loading. This lends the system to optimal positioning of the tank pressure for ship loading as distinct to steady state operation in between ship loading.
APLNG site background. ConocoPhillips operates an LNG facility on Curtis Island, Queensland, Australia on behalf of Australia Pacific LNG (APLNG). ConocoPhillips has successfully applied APC to the two liquefaction trains to deliver operability and economic benefits. Recently, the site turned attention to the development of a BOG APC application to improve the operability and optimization of this part of the process, which has a profound effect on the liquefaction train’s stability and optimization.
At APLNG, the process design is based upon pure component refrigeration circuits driving the liquefaction process, with the LNG storage tank BOG routed back into this process. Accordingly, the primary optimization objectives were to:
FIG. 1 outlines the nature of the APC system with some duplication deliberately omitted to simplify the diagram. This includes the following:
Compressor control programmable logic controller (PLC) interface. The site is equipped with three BOG compressors of which two are online at any one time. The BOG compressor control and anti-surge protection systems are hosted on a third-party PLC—in effect, a separate process control island from the main process DCS. With some of the original construction project scope not executed, the nature of the PLC control left some optimization opportunity unrealized, including the following:
While some sites with similar BOG control designs have chosen to eliminate the interaction by staggering the suction pressure setpoints (in effect, scheduling the loads on the compressors), this results in an efficiency loss through the net recycle flow being greater than necessary for the process needs. At the APLNG site, the operators were pursuing a compressor balancing objective by making regular changes on the suction control setpoints to try and balance the operation of the two machines (trading operator workload for inefficiency).
FIG. 2 illustrates the difference in operation sought by coordinating the IGV and RV positioning via the APC—improving efficiency by minimizing the amount of time that additional power was consumed with incremental IGV opening to support additional RV opening. That is, the operators ideally wanted to see either the IGVs or the RVs open but not both simultaneously.
While the natural instinct was to collaborate with the PLC control system vendor, travel restrictions during the height of the COVID pandemic and support being on the opposite side of the globe encouraged the approach of developing the APC interface to the PLC control locally. Some initial exploratory step testing using the existing suction pressure controls proved that maintaining these loops as the underlying basis of the APC was not going to be successful. Fortunately, it was feasible to develop a means of getting direct control of the IGV and RV on each compressor while retaining the high frequency anti-surge protection provided by the PLC control.
APC design challenges. Although the APC design is simple on the surface, there are several unique design aspects that are unusual:
APC implementation challenges. Despite the relatively small dimensions of the APC model matrix (i.e., relatively few inputs and outputs), there were several challenges associated with the implementation of this APC application:
This is a significant sequence of events to achieve in a single campaign and a testament to the commitment of the project team and the cooperation offered by the site operators. The project team onsite were able to resolve the challenges associated with the tight timeframe and achieve a good result, which enabled the new APC application to stay online with some minimal tweaks to the shiploading sequence completed over the following few months. Some other minor modifications were made to accommodate single BOG compressor operation to improve the APC uptime during single train operations. Further changes to the shiploading sequence are expected to result from improvements in the manual procedure that are planned.
Project benefits. The outcome of this APC development exercise was very satisfying for the site with a range of benefits realized:
To quantify these benefits, the site’s liquefaction APC model gains were used to determine that the reduced BOG flow resulted in a 0.25% production capacity increase for the facility. If the project and ongoing maintenance costs associated with the application are considered, the project had a payback period of less than 1 month and an extremely favorable net present value determination for the APLNG JV.
In conclusion, this APC development is an important step forward in the site’s uptake of APC technology, proving that the benefit:cost ratio of additional applications can be significant once the technology is established onsite through higher absolute value foundation applications. This APC project outcome is an excellent example of how both the development process and the final APC application can achieve the important objectives of:
ACKNOWLEDGEMENTS
The authors would like to thank Australia Pacific LNG (APLNG) and its shareholders (ConocoPhillips Company 47.5%, Origin Energy Ltd 27.5%, and Sinopec 25%) for their support of this improvement project and their permission to publish this article.
ANDREW TAYLOR is a Director/Principal Consultant at Greenfern Dynamics. He has 30 yr of experience with APC in various industries, including oil refining, LNG production, oil and gas, chemical production and minerals processing. He earned a BE degree in engineering science from the University of Auckland and is a chartered professional member of Engineers Australia.
TOM WILSON is a Senior Process Control Engineer at ConocoPhillips Australia. He has 14 yr of experience in the oil and gas industry. He earned a BE degree in mechatronic engineering from the University of Queensland and is a chartered professional member of Engineers Australia.
ROBERT DAVIS is a Process Engineer working for ConocoPhillips Australia. He has 7 yr of experience in the oil and gas industry working with the Optimized Cascade Process. He earned a BE degree in chemical engineering and is a chartered member of IChemE.
FRANCOIS PIETERSE is a Consultant at Greenfern Dynamics. He has 15 yr of experience with APC in various industries, including oil and gas, minerals processing, and pulp and paper. He earned BE degrees in chemical engineering and process control from the University of Pretoria and is a chartered professional member of Engineers Australia.