Developing standards in refinery 
energy management

Precise validation and reconciliation of site data is required for an effective energy management system in a refinery

Hervé Closon and Ulrika Wising, Belsim
Jean-Claude Noisier, Société Ivoirienne de Raffinage

Viewed : 6292

Article Summary

Energy efficiency and the cost of energy are top priorities for heavy energy consumers. For refineries, energy is often the single largest component of variable operating expenses. The implement-ation of an energy management system for existing energy assets is 
an alternative that requires less investment and can often be implemented quicker than traditional methods for reducing energy costs.

Energy management seeks to influence how energy is managed in industry, which is the single greatest barrier to realising potential large-scale improvements in energy efficiency over the coming years. The challenge is to provide a business-friendly mechanism for applying the same types of management practices to energy as are already applied 
to other resources such as labour 
and materials.

Energy management systems
There is a movement today to develop a standard for energy management in industry, much like the standards for quality management. This standard will most likely cover energy efficiency, energy performance, energy supply, procurement practices for energy using equipment and systems, and energy use. The standard under development will also address measurement of current energy usage, and the implementation of a measurement system to document, report and validate continual improvement in energy management.

Features of a standard for an energy management system may include:
•  An energy policy that defines scope, objectives and targets for the energy management system and addresses all significant energy use
•  A strategic plan that requires measurement, management and documentation for continuous improvement in energy performance and efficiency
•  A cross-divisional energy manage-ment team led by an energy coordinator, who reports directly to refinery management and is responsible for overseeing the implementation of the strategic plan
•  Policies and procedures to address all aspects of energy purchase, use and disposal
•  The identification of key performance indicators, unique to the company, which are tracked to measure progress
•  Projects to demonstrate continuous improvement in energy efficiency
•  The creation of an energy manual, a living document that evolves over time as additional energy-saving projects and policies are undertaken and documented
•  Periodic reporting of progress to management based on these measurements.

In a nutshell, energy management systems offer an expert and 
best practice-based framework for organisations and enterprises to develop goals for energy efficiency, to plan interventions, to prioritise energy efficiency measures and investments, to monitor and document results, and to ensure continuity and constant improvement in energy efficiency.

In order for an energy management system to function well, it needs to be fed with numerous data and information. One of the biggest challenges for an energy management system is the quality of the process data it depends on, which may be biased by errors, including:
•  Intrinsic errors linked with sensor technology
•  Incorrect calibration
•  Poor location of sensors
•  Measurement noise.

One way of assuring the quality of process data in an energy management system is to include an advanced data validation and reconciliation step in such a system.
Data validation and reconciliation
Data validation and reconciliation (DVR) are two rather young sciences  developed in the late 1950s. They are different technologies that can be used together or separately. There are also different approaches to the two technologies that define their rigour. In order to guarantee data quality, a combined approach using the most rigorous approaches to the two technologies is recommended. This is referred to in this article as advanced data validation and reconciliation.

Data validation is the process of ensuring that a program operates on clean and correct data. It uses routines that check the correctness and meaningfulness of data. The most common method used for data validation is the range check, which checks that the data lay within a specified range of values. More advanced approaches use different algorithms to identify gross errors and eliminate them. Validation, by definition, does not close mass, energy or component balances and the range check is based on experience, which 
is itself based on the observation of 
raw measurements.

Data reconciliation is the process of adjusting data in order to close mass, energy and component balances. Most approaches to reconciliation do not, for example, check whether the provided values are included in the range of feasible values. Nor do they include mass, component or thermodynamic constraints that can determine whether flow rates are negative, or find unfeasible temperatures.

Limitations of data validation and/or reconciliation
There are different levels of rigour to apply to data validation and data reconciliation. When using simple approaches such as a basic range check for validation, without taking thermodynamic constraints into account for reconciliation, the solution is not a robust and complete solution answering to the need for accurate process data. Such basic data validation and/or reconciliation (basic DVR) systems are limited, especially when validation and reconciliation are used separately.

Basic DVR systems usually employ a linear equation solver for mass balance and heat exchanger balance calculations; they do not apply a non-linear equation solver simultaneously handling mass and complex thermodynamic balances such as phase equilibrium. Basic DVR systems do not have embedded thermo-dynamic databases and it is necessary to convert measurements (temp-eratures, pressures) using enthalpy correlations prior to any reconciliation of their energy balances. This sequence is done outside the basic DVR application and does not increase the redundancy of the model. As a result, the reliability and accuracy of the results are limited, especially for an energy management system.

Add your rating:

Current Rating: 2

Your rate:

  • Responsive image 2022 Opportunity Crudes Conference 24-26 October
  • Responsive image Neles is now part of Valmet
  • Responsive image PTQ - Read The Latest Issue
  • Responsive image FCC online diagnostic services for FCCU
  • Responsive image Extensive tray portfolio
  • Responsive image Shut-off valves
  • Responsive image Nasty stuff
  • Responsive image ViscoSure: In-line Viscosity
  • Responsive image FluegasExact 2700 effective combustion analysis
  • Responsive image Catalysts for optimal performance