Aug-2021

Improve energy efficiency while reducing CO2 emissions sustainably

A solution that captures real-time process data can optimise furnace operations, energy losses and carbon emissions into the atmosphere.

Avnish Kumar
LivNSense Technologies Private Limited

Article Summary

The petrochemical industry accounts for 6% of the energy usage in the US. Roughly half of that comes primarily from oil and natural gas, and it is estimated that 60% of a plant’s energy consumption comes from furnace operations. This article focuses on the optimisation of furnace operations to achieve greenhouse (GHG) emissions reduction. 

While most furnaces are designed for a thermal efficiency of 70-90%, actual operating efficiencies are much lower. Plant managers are constantly challenged by operating conditions and the need for continuous optimisation due to:

- Deterioration in throughput over the furnace’s lifetime
- Increased wastage of consumables and raw materials
- Unpredictable and longer downtimes, leading to lost production and increased costs
- Increased energy losses and carbon emissions

As a furnace ages, these challenges exacerbate, leading to early replacement. While modern furnaces have electronics that gather data, there is little or no ability to perform real-time analytics and predict events in advance or provide predictive insights into interventions needed for optimal operation.
A typical 500 KTA capacity ethylene plant consumes 30 MW hours of electrical energy per year. Even a 1% reduction in energy consumption from furnace operations will reduce millions of tonnes of emissions into the atmosphere (equivalent savings of £5 billion/year).

With the World Economic Forum (WEF) driving ‘The Net-Zero Challenge’ to reduce carbon emissions globally, LivNSense is driving the future of energy efficiency through its Cognitive Funace4.0 platform. This enables an innovative solution, CarbonSense, to improve energy efficiency while reducing CO2 emissions sustainably and cost-effectively.

Understanding furnace operations
Combustion sources such as furnaces play a critical role in the process industry and require large amounts of fuel (gas, fuel oil). As a result, combustion efficiency directly influences the performance and operational costs of production facilities. Also, furnaces do not constantly operate under the design conditions, which is another major cause of efficiency issues.

However, efficiency is not the only concern. Compliance with emission standards and safety are significant challenges too. Incomplete combustion occurs when insufficient excess air is supplied to burn all the fuel completely. As a result, large amounts of CO and H2 are formed, making the burner extremely inefficient. This reduces the flame temperature and may encourage the operator to increase the flow of fuel, worsening the carbonisation. Hence, the measurement at the arch determines whether complete or incomplete combustion takes place at the burner level.

Business process understanding
The oxygen level inside the furnace is measured at three different places: the stack, arch and SCR. If excess oxygen is present in the arch O2 sensor, an excess amount of oxygen is being used in the furnace.

Typical ranges of oxygen at various locations in the furnace are:
          Equipment level         % oxygen present
          Arch                           2-3
          SCR                           3-3.5
          Stack                         3-4

It is important to measure the oxygen level in the arch because combustion of the fuel gases takes place prior to the arch. By measuring the oxygen level here, we can determine whether the reaction inside the furnace is complete or incomplete combustion.

CarbonSense solution
CarbonSense achieves a reduction in energy consumption and GHG emissions and an increase in furnace run length by optimising the furnace operating parameters of petrochemical and chemical plants.

In a typical ethylene production furnace, coke formation is a major problem that reduces heat exchange and leads to more fuel being burnt, and thus more emissions being released into the atmosphere. Coke formation also leads to reduced furnace run length. Addressing fuel composition changes by leveraging the real-time thermodynamics model and AI/ML techniques enables continuous bias of the air/fuel ratio to stabilise combustion and heat transfer into the tubes.

To address these issues, we have leveraged the Cognitive Furnace4.0 platform for data acquisition from DCS/historians in real-time, pre-processing and modelling based on first principles and machine learning algorithms. LivNSense has also levered its unique IP - disturbance index for estimating manufacturing condition variations in real-time. The outcome is delivered through live furnace digital twins for continuous improvement and high accuracy. Our holistic platform has also been proven to increase furnace efficiency and productivity, as well as reduce NOx and CO2 emissions.

A Process-AI Platform with out-of-the-box features:  
- Core AI process analytics module: Module for data pre-processing suitable for major process industries for statistical analysis.
- AI/ML custom module: AI models are built using industrial-grade data simulated from plant operations, models are strengthened by first principles and advanced analytical approaches trained for the specific process industry.  
- Digital twins: Cognitive Furnace4.0 platform (iSense4i) uses a digital twin, a virtual representation that matches the attributes and operational metrics of a ‘physical’ production line through the captured production-line data.
- SMARTEDGE AI-based Edge firmware for connectivity, device and security management, deep learning and AI vision at the Edge for real-time OT-IT integration. Ability to support millions of events in real-time on custom and third-party HW platforms.
- The solution is platform agnostic and can be deployed on-site, hybrid cloud and public cloud environments. The solut