Modeling the Smells

Guidelines for the proper development of a modeling system for industrial odors.

Measuring and modeling the scope and impact of odor emissions is a task that should be assumed by any industrial activity that generates them. First, to avoid causing environmental problems or damage to the health and quality of life of the community surrounding its facilities, conditions that companies must meet to achieve sustainability. And then, because soon in Chile several items will have to respond to the regulations that are being developed, starting with the pig-raising sector, whose regulation –until the closing of this edition– was in the process of being approved by the Comptroller General of the Republic.

In fact, said provision indicates that the regulated must carry out measurements and model their odoriferous emissions and then report the results to the Superintendence of the Environment (SMA) that will act as a supervisory entity.

On this scenario, below, we provide some guidelines and recommendations to properly develop an industrial odor modeling system.

Technical Guides

The modeling of atmospheric pollutants, among which are odors, is a tool that aims to quantify the concentration levels of pollutants in the receptors and determine their scope, based on meteorological records in real time. In this way, it seeks to anticipate possible scenarios in which the quality of air and life of the nearby population may be affected, with a view to adopting timely measures to avoid these risks.

In recent years, in our country, environmental institutions have developed documents that contain useful technical information so that companies can carry out odor modeling that helps predict the impacts of emissions.

Daniela Caimanque, a professional from the Department of Noise, Light and Odors of the Ministry of the Environment (MMA), details: "The reference used for the modeling of odors is the 'Guide for the Use of Air Quality Models in the Evaluation System of Environmental Impact (SEIA)', whose purpose is to establish criteria for the use of air quality models in the SEIA as an estimation tool. Additionally, the 'Guide for prediction and evaluation of odor impacts in the SEIA' is available, whose objective is to provide tools to estimate odor emissions, as well as predict and evaluate environmental impacts. Both guides were prepared by the Environmental Assessment Service".

The specialist also announces that as part of the Strategy for Odor Management in Chile and the future odor regulations that will come into force, the Ministry of the Environment plans to prepare a technical document that complements the aforementioned Guides, specifically for the use of odor models. This will be done, "considering the particularities of this pollutant, the limitations and uncertainty inherent in each type of model, as well as the uncertainty associated with the data used as input information to the models. In this way, the Ministry hopes to deliver greater guidelines for project holders to evaluate the results of odor modeling within the framework of environmental management instruments related to this contaminant," says Caimanque.

Stages of development

Beyond the guidelines that the aforementioned guides or other technical documents can provide, practical experience provides a series of considerations to take into account.

On this scenario, from TSG Environmental, a company specialized in the development of odor management solutions, they first explain that, in general, the implementation of a real-time modeling system has two stages: odor diagnosis and pre-engineering .

"The first stage contemplates the quantification of odor emissions from the process units (emissions inventory), either based on an odor impact study or a theoretical estimate of odor emissions, which allows identifying and characterizing the main sources of odor emission that will be entered into the system for online monitoring", they detail.

The next phase contemplates the pre-engineering of the industrial installation. Here, initially "the computational requirements necessary to achieve optimum performance of the system are evaluated, based on the magnitude of the industrial activity and the complexity of the geographical environment where the emissions will be projected. Once defined, the system is configured to represent local meteorological conditions and emissions dispersion, to then give way to the entry of the emitting units to the system considering structural characteristics, flow and odor emission", they indicate.

After that, the start-up of the system begins, a period in which an exhaustive quality control is carried out, evaluating the response times, stability, performance and coherence of the projected odor plumes.

The development and implementation process of these systems is not exempt from technical difficulties that must be addressed to ensure correct operation.

Which are the most important?

In TSG they answer: "The main challenge in this type of real-time systems is related to the construction of an adequate meteorological base on the complexity of the terrain, that is, with the ability to represent the local dispersion conditions around the industrial installation. ". They add that this requires extensive technical knowledge in modeling (configuration, management of models and input variables, among other aspects), which makes it possible to ensure the quality and representativeness of projected emissions in real time. These factors are key to the stability of the system, with a view to avoiding numerical inconsistencies.

Specialists also warn that it is essential "to have quality in the input data of the system for the characterization of emissions associated with the sources of industrial activity. This is related to the representation of operational cycles and communication systems in the case to integrate continuous emissions monitoring (CEMS) The sensitivity of the system requires periodic and comprehensive review of the input and output data Due to the magnitude of the information flow, it is necessary to implement intelligent and automated processes for an exhaustive quality control of the information, since an erroneous characterization of the real projection of emissions could lead to inaccurate and/or erroneous results that hinder an adequate management of the resources focused on minimizing the impacts on the communities".

Technology and Big Data

In general terms, the development of odor modeling systems requires various technologies and computer tools to manage and process the information that comes from the monitoring of emissions and the meteorological or other variables that are entered into the system.

By way of example, from TSG they comment: "Our systems are based on servers hosted in the cloud whose multiprocessor technology allows us to achieve high performance in the mathematical and predictive processing of complex models, validated and approved by environmental protection agencies, with which our real-time modeling system operates".

They add that the high flow of information associated with the continuous registration and storage of data has led them to implement Big Data tools for the management and analysis of information, in order to optimize response times and improve the user experience in use. of the different analysis and operational control platforms. "In addition, our system uses among its components high-resolution global forecast weather information from the US National Center for Environmental Prediction (NCEP). This information is adjusted locally by incorporating local weather stations that They allow to increase the representativeness, complemented with Big Data analysis tools to understand the dynamics of meteorological events when evaluating complaints by the community", they point out.

What could be the estimated range of investment in these solutions? "The price depends on the number of source groups to model, the contracted functionalities and the number of receivers to cover. To indicate an order of magnitude, it can be considered that the implementation, enabling and maintenance of the platform and its tools are is in the range of 35,000 to 50,000 USD per year", they answer in TSG.

Accuracy and Validation

By the way, any odor modeling system aims to deliver results that are as accurate as possible. How can this be achieved and the information they deliver validated?

In TSG they state: "To achieve a high degree of precision in this type of system, it is essential to have representative and quality information, that is to say that the input data is obtained through standardized methodologies and validated by organizations with technical competence, both In the same way, it is of the utmost importance that the information associated with the process units of the industrial installation be adjusted in the best possible way to the real conditions of operation, both in structural terms and in the characterization of emissions and their operational cycles".

They warn that when continuous pollutant monitors are integrated into the real-time modeling system, it is essential to apply calibration and maintenance programs for the different sensors associated with the recording of variables. And they point out that the information can be contrasted with measurements from air quality stations located near the industrial facility.

They add that, in the case of odor modeling, the results are generally compared with the records of community complaints where odor notes associated with the process units of the industrial facility are identified.