Natural gas, the cleanest of the fossil fuels, is the fastest growing primary energy source for the world today. Consumption of natural gas is projected to increase by nearly 70 percent between 2002 and 2025 [1]. Liquefied natural gas (LNG) is the most economical way to transport natural gas over long distances. Reductions in costs throughout the LNG chain, advances in LNG technology, etc. have transformed LNG into an increasingly global energy option similar to oil. In only one quarter of a century, the international energy scene has witnessed a remarkable growth in LNG trade. As an alternate fuel, the demand of LNG is doubling every ten years. In 2001, world's total LNG demand was estimated to be over 100 mtpa [2]. By 2012, it is expected to be 270 mtpa. With the expectation of increasing demand for energy with time, LNG has established itself as the fuel for the future. In 2005, the global liquefaction capacity for LNG was 150 mtpa [3].

An LNG plant is highly energy-intensive and refrigeration section is the main consumer of energy. The operational flexibility and efficiency of the refrigeration section are critical to the overall efficiency of the LNG plant. Main cryogenic heat exchanger or MCHE is the heart of the refrigeration section and is the most important heat-transfer equipment in a base-load LNG plant. It cools and liquefies natural gas to –162 C. For process optimization, we need to look from an integrated point of view and MCHE is one of the most critical equipment in LNG process. Modeling MCHE is a crucial first step in the optimization of an LNG process. Even a relatively small improvement in MCHE or other part of refrigeration section can have a considerable effect on the entire process and its operating costs.

MCHE is normally a multi-stream spiral-wound heat exchanger. Its special features include multiple hot/cold streams, partial direct heat transfer via mixing of streams, stream splitting, high density of heat transfer area, etc. These permit large heat transfer at small temperature differences and make this type of heat exchanger extremely popular in LNG plants. However, little work exists on the modeling of MCHE in the open literature. In principle, one could model it in two ways. One is to use rigorous computational fluid dynamic models, but this can be prohibitively time-consuming in an optimization study, and details of MCHE design are largely proprietary. An alternate approach would be to develop simpler, approximate models that could predict the performance of an existing MCHE without knowing its physical details. This is what we plan to do in this paper.

Our objective is to address the rating of a MCHE or the prediction of its performance based on historic data. Yee et al. (1990) presented a model for simultaneous targeting of energy and area for heat exchanger network synthesis using the idea of superstructure. However, they addressed the design problem for a general network with utilities rather than an operational problem for a multi-stream exchanger. In this paper, we employ the concept of superstructure representation for heat integration and heat exchanger network synthesis to model a MCHE and use real plant data to derive an equivalent 2-stream exchanger network. Thus, we model MCHE simply as a superstructure comprised of two-stream exchangers, where mixing and splitting of streams (to find all the possible matches between hot and cold streams) are allowed. In other words, the objective is to fit the model with plant data for an existing spiral-wound heat exchanger rather than finding the area or minimizing the cost. We will use a mixed-integer nonlinear programming approach to select the best network that describes an existing MCHE.

We will validate the proposed model using historic plant data from an actual main cryogenic heat exchanger. If possible, we will also assess the ability of our model to predict the performance and its usefulness in plant operation.

References

[1] International Energy Outlook 2005, July 2005

[2] B. P. Chedigaz, Statistical Review of World Energy, June 2002.

[3] Global LNG Market Worldwide (2005-2015), Research Report, January 2006.

[4] Yee, T. F.; Grossmann, I. E.; Kravanja, Z. Simultaneous Optimization Models for Heat Integration-I. Area and Energy Targeting and Modeling of Multi-Stream Exchangers. Comput. Chem. Eng. 1990, 14 (10), 1151-1164.