Evaluating the design of industrial systems for cost-effective decarbonisation
Tyndall Research Theme 2 - Decarbonising Modern Societies
Project ID - TRS/05
Technical Summary
The objectives of this project are:
1. To develop a new methodology to design industrial utility systems, such that they reduce greenhouse gas emissions in the most efficient and economic way. Thus, the methodology has to combine the traditional process optimisation techniques with the strategy of distributed optimisation in order to tackle large-scale systems with acceptable precision. It is necessary to incorporate into the methodology the consideration of the options for greenhouse gases minimisation, so that this minimisation leads to minimum cost penalty, or even a profit.
2. To implement the methodology in a computer-based modelling framework, together with the appropriate procedures for industrial data extraction and design directions for the industry engineers.
3. To apply the methodology on a real industrial case provided by the collaborating organisation. This will also include validation and procedural consistency improvement of the methodology.
4. To develop scenarios based upon stages (1), (2) and (3) which will extrapolate the findings of the research to higher scales (e.g. to that of a national economy assuming different levels of implementation of the technologies investigated). The research can thereby contribute to the Tyndall Flagship Project on 'Transition to a Decarbonised UK' by providing inputs on what levels of carbon abatement can be realistically achieved from re-design of industrial systems.
At the Department of Process Integration - UMIST a series of energy saving methodologies have been developed.
In the early 1980-s the Pinch Technology for Heat Exchanger Networks (HEN) design was developed (Linnhoff 1983). The methodologies evolved to incorporate total cost targeting (Linnhoff and Ahmad 1990, Ahmad, Linnhoff and Smith 1990) and block-decomposition based HEN synthesis. In parallel with this, a HEN retrofit framework, based on the "process pinch" (Tjoe and Linnhoff 1986) and "network pinch" (Asante and Zhu 1996) concepts has been established.
In the context of the total site, consisting of a number of process plants, the utility system has also been a subject of intensive research. (Dhole and Linnhoff 1992) have given an energy-targeting framework, based on the total site profiles. Later on, methodologies for combined heat and power co-generation by the site utility system have been developed - T/H model (Raissi 1994), subsequently improved (Mavromatis and Kokossis 1998-a and 1998-b).
Based on the enumerated targeting methods, Mavromatis and Kokossis (1998-b) developed a steam-turbine network synthesis methodology, which deals with the steam distribution part of the utility system. Also, the methodology assumes fixed number and pressure levels of steam distribution headers, choosing amoung several pre-specified alternative steam header sets.
A research work by H.Singh (1997) provides a methodology to design utility systems for minimum emissions. The study puts an emphasis on the ways for assessing the cost, associated with achieving the environmental regulations at the industrial sites. This methodology relies on pre-specified sets of steam levels and on a single superstructure optimisation.
A very important parameter used in the design, retrofit and operational analysis of utility systems is the site power-to-heat ratio referred to as R-ratio. The R-curve is a plot of cogeneration efficiency of the utility system versus site power to heat ratio (R-ratio). It represents the site utility system, showing the variation of cogeneration efficiency over a wide range of R-ratios. The application of the R-curve concept to utility systems (Kimura and Zhu, 2000), allowed for consideration of the co-generation efficiency of the sites. This gives a useful screening tool to compare design alternatives. However, the comparison on the basis of the thermal efficiency is not always sufficient, as high thermal efficiency always comes at certain capital cost and, therefore, does not always mean high economic efficiency.
Finally, a hierarchical, two-level methodology to design power plants has been presented by Manninen and Zhu (2000). This work deals with the design of commercial power stations. It provides interesting directions for considering the strategy of thermal systems synthesis. The design framework defines two stages and respectively - two models of the power plant: master (high-level) model and conceptual (more detailed) model. In this framework, the first stage, effectively optimises system features less sensitive to details (like presence of a gas turbine and HRSG, boiler firing) by using a simplified power plant model. Later on, in the second optimisation level, a detailed model is employed. It is used to optimise the detailed design of the plant, considering actual equipment options.
A common feature of the enumerated methodologies is, that all they deal with the problems of different parts and different detailed contexts rather than the overall problem of the synthesis and design of the total site energy system. As such, they represent a useful basis, or bricks, for development of a global methodology for designing utility systems. The methodologies of H.Singh (1997) and J.Manninen (2000) rely heavily on integrated superstructures, including all design features likely to be applied. This imposes large and complex sets of design equations to be put simultaneously in single optimisation model formulations. The latter methodology (Manninen, 2000) introduces a hierarchical decomposition in terms of the algorithm implementation.
A major ambition of this thesis is to adapt the above methodologies for the explicit consideration of decarbonisation, whilst taking into account the other specification requirements for utility systems. This has not yet been achieved, yet is an important step in providing useful knowledge to highly carbon-intensive operators. The Intergovernmental Panel on Climate Change (IPCC) performs regular assessments of the policy-relevant scientific, technical and socio-economic dimensions of climate change and has considered technologies available for carbon mitigation in its Second and Third Assessment Reports. These assessments are, however, typically highly generic and refer to the global scale without treating individual sectors and processes in detail. Such assessment, whilst important, does not meet the needs of many industrial sectors, for which a more detailed analysis, based upon technical models of facilities and operations, is required. The proposed work can be seen as a form of 'ground -truthing' of the IPCC and similar global assessments in more specific detailed circumstances. This more realistic assessment in particular carbon-intensive sectors of industrialised economies is likely to be more credible to the industry.
It is, furthermore, possible to use the technical tools here developed to explore various policy scenarios for decarbonisation that are currently being implemented or debated by European and other nations. For example, the relative effectiveness and costs of carbon taxation, carbon emission limitations, carbon emissions trading and energy efficiency agreements can all be analysed with the technical model that will be developed. Individual company policies can also be assessed, such as British Petroleum and Shell Group targets' to reduce their emission levels by 10% compared to the levels of 1990 by 2010 and 2002 respectively. The effect of other exogenous changes, such as rising fuel prices, can also be studied. The results of different policy scenarios will be extrapolated from the individual operation scale to the UK national scale using different levels of implementation of technologies (assuming different discount rates, taxation rates, energy prices, etc.). The change in carbon emissions at the national scale can then be estimated, for inclusion in UK-wide decarbonisation analysis to be undertaken by Tyndall.
Based on the analysis presented above, the current research project has the following stages and milestones.
METHODOLOGY DEVELOPMENT
(1)
To introduce further decomposition in the general design algorithm, leading to a three-level procedure as follows. The topmost level uses the task formulation as input. The other two levels use as input the information, generated at the previous level.
Level 1: High-level screening model. It is intended to take the initial task formulation and to produce the following outputs. A rough total utility system cost estimate, suggestions with respect to steam generation blocks to be employed - boilers, gas turbines with HRSGs (Heat Recovery Steam Generators), or both options, a range for the steam level number, which is likely to be most profitable.
Level 2: Middle-level model. This model has to use the available heat and utility targeting techniques (Linnhoff 1983, Ahmad 1990, Mavromatis 1998, Shang 2000) in a distributed optimisation framework in order to further clarify the system capital and operating cost and to optimise the number and exact levels of the steam headers.
Level 3: Detailed model. This model has to employ rigorous equations and practical equipment options for the final utility system design. As a result the final optimal utility system design will be produced.
If any lower-level model reveals facts and details in substantial disagreement with the models from the upper levels, a feedback mechanism and looping has to be applied.
(2) To introduce in the middle-level and the detailed-level models structural and functional decomposition of the total industrial site with respect to the utility system. This will result in the decomposition of the single complex model into a set of simpler utility and process subsystem models, thus allowing more precise modelling in terms of precise equations and detailed design of the subsystems. This decomposition has to be implemented with distributed mathematical programming model formulations.
(3) A special emphasis is put on the environmental impact of the industrial site. The methodology has to combine the consideration of the classic filtering end-of-pipe facilities with technologies of reduced pollutant production such as low-NOx burners. Another direction of the research will be the assessment of the economic feasibility of the newly emerging techniques for CO2 reduction such as fuel switching and carbon dioxide fixation with bio-scrubbers, pressure swing adsorption for CO2 capture and recovery.
(4) The systematic optimisation of all the options: for efficiency improvement, pollutant production minimisation and end-of-pipe filtration has to produce designs, achieving emission regulations in an economic way, with a minimum cost penalty and/or maximum profit.
INDUSTRIAL MANAGEMENT, FEEDBACK AND TECHNOLOGY TRANSFER
In order to make the developed methodology practical and to validate some of the subsystem models, an industrial feedback is vital. In order to achieve this, the following activities will be performed.
(5) Data collection sessions. Two types of sessions are provided. Firstly, acquiring experimental data to validate the models of different equipment items. Secondly, investigation of extreme emergency what-if scenarios. The latter will ensure the consideration of proper functioning of industrial sites within the environmental limits in emergency situations.
(6) Incorporation of the project management issues and cash flow planning into the design framework. This will ensure the practical applicability of the methodology.
(7) Exploration of different medium to long-term scenarios of fuel and power prices change, and different carbon mitigation policies (taxation, trading, regulation, etc.) in order to estimate the cost implications on future industrial projects prior to the start of the actual building works. Medium and long-term market forecasts for trends of the power from electricity grid and the fuel prices will be taken and divided into operating periods. The so derived periods will be accounted for by the economic evaluation of the industrial sites.
(8) Extrapolation of the findings through scenario analysis to the UK national scale.
References
B Linnhoff and E Hindmarsh (1983), The Pinch Design Method of Heat Exchanger Networks, Chemical Engineering Science, Vol.38, No 5, pp 745-763
B Linnhoff and S Ahmad (1990), Cost Optimum Heat Exchanger Networks-1. Minimum Energy and Capital Using Simple Models for Capital Cost, Computers Chem Engng, Vol. 14, No 7, pp 729-750S
Ahmad, B Linnhoff and R Smith (1990), Targets and Design for Detailed Capital Cost Models, Computers and Chem Engng, Vol. 14, No 7, pp 751-767
T N Tjoe and B Linnhoff (1986), Using Pinch Technology for Process Retrofit, Chemical Engineering, pp 47-60, April 28
N D K Asante and X X Zhu (1996), An Automated Approach for Heat Exchanger Retrofit Featuring Minimal Topology Modifications, Computers Chem Engng, Vol. 20, Suppl. pp S7-S12
V R Dhole and B Linnhoff (1992), Total Site Targets for Fuel, Co-generation, Emissions, and Cooling, Computers Chem Engng, Vol. 17 Suppl. pp s101-s109
K.Raissi (1994), Total Site Integration, Ph.D. thesis, DPI-UMIST
S.P.Mavromatis and A.C.Kokossis (1998-a), Conceptual Optimisation of Utility Networks for Operational variations-I. Targets and Level Optimisation, Chem.Eng.Sci. 53, 8: 1585-1608
This research is supported through a Tyndall Research Studentship.
Student: Petar Varbanov
Institution: Department of Process Integration, UMIST
Lead Supervisor: Dr Jiri Klemes
Dates: April 2002 - July 2003
Collaborating
organisation: Process Integration Research Consortium
Printer-friendly version
Send by email