United Nations Development Program Kuwait Environment Public Authority

United Nations Development Program Kuwait Environment Public Authority

Oct. 1, 2022
Juan Guzman
Published: Oct. 1, 2022

In 2012, the Kuwait Environment Public Authority (KEPA) issued new air quality regulations to categorise sources which cause, or contribute significantly, to air pollution. As part of the regulations, KEPA also created the Kuwait Implementation Plan (KIP) to ensure compliance with Kuwait Ambient Air Quality Standards (KAAQS). As part of the KIP, Kuwait was divided into Air Quality Control Zones (AQCZ) based on monitored ambient air quality.

Creating a benchmark inventory using Palisade’s risk analysis toolset

Consultants from United Nations Development Program (UNDP) assisted KEPA to create a national emissions inventory to provide a countrywide air modeling capability that would allow a source permit and monitoring programme. At the outset of the project, UNDP found there was a vast amount of organisations within the region with industrial processes that created air emissions, but that recordkeeping was unusable. In many cases, required parameters such as atmospheric pressure and flow rates were not recorded. To overcome this lack of data, UNDP turned to Palisade’s risk analysis software @RISK to capture the full range of possible outcomes as opposed to establishing one particular static number. Palisade’s risk analysis software gave UNDP the ability to justify the results of its research to KEPA by showing the range of results. This allowed KEPA to understand what data they really needed in order to ensure correct measurement of emissions.

Helping Kuwaiti businesses understand their contribution to air quality

Understanding contribution to air quality is complex. Wind conditions and production output from industrial sources can contribute to a high concentration of pollution in an area. When the weather changes, the concentration changes as a result. Brian Freeman, Team Leader for Air Regulatory Management Systems at UNDP, identified @RISK as an integral part of facilitating the success of the project. “As part of the analysis, we needed to provide input parameters into complex air dispersion models. These parameters included physical dimensions such as stack heights and diameters that were easy to identify. The more important parameters were independent variables such as release rates, temperature, and flow rates. This was where Palisade’s @RISK was particularly helpful. We were able to model the worst-case to best-case scenarios for each parameter in our models to show long-term exposure, as well as contributions from individual sources.”

Using Monte Carlo Simulation is fundamental to success

In this ongoing project, UNDP employs @RISK to correlate local industries’ impact on air quality in Kuwait. Using Monte Carlo Simulation, UNDP looks at potential ranges of impacts. The analyses can show quite conclusively, for instance, that a particular area with specific companies may not be the sole contributors to poor air quality, and are contributing significantly less than was initially considered. Two different results graphs offer a worst-case, with a particular factory contributing 4% on a bad day. But 95% of the time, based on wind patterns, that factory’s contribution was less than 2%. UNDP was also able to take years of historical monitoring data and fit the information into distribution models. The resulting graphs were output as PDFs, and allowed for year-by-year comparisons to show improvement or deterioration of pollutant concentrations. Comparative statistics were incorporated to determine overall improvement by using the mode and the 95th percentile.

Brian Freeman reports that Palisade’s software “is an essential tool that is used daily in our organisation.”

"The ongoing work with KEPA has meant Palisade’s products are a very integral part of the success of this programme. Pretty much if we don’t use @RISK at least once a day, it means we’re on vacation."

Brian Freeman
Team Leader for Air Regulatory Management Systems, United National Development Program

Favoured features and distributions

Binomial distribution – UNDP uses binomial distributions as switches – P1 representing one condition, P2 representing another condition. In many instances they see data from the same source follow one distribution over one set of conditions and follow another distribution over another set of conditions. They might not know what those input conditions are but the results are clear when shown on a histogram. Using a binomial distribution they can assign a probability that Condition P1 might happen and a probability that Condition P2 might happen. If Condition P1 happens, they use PDF 1, if P2 happens, they use PDF 2. The following example showing measured Sulfur Dioxide (SO2) shows how UNDP used the binomial distribution to create a composite PDF.

Triangle distribution – UNDP uses the triangle distribution when the underlying distribution of a data set is not known, however it has finite boundaries. It is very good for first-draft models when little is known about the system and limited data points are available.

Results graphs demonstrate significant changes – Below is a model demonstrating the overall yearly improvement or deterioration of ambient air. In this case, overall SO2 has decreased by 15%.

Freeman concludes, “I’ve used @RISK within my work for over 10 years. Nothing else matches the flexibility and ability of the software. Other products lack the interface ability completely, and whilst MCS is useful, it’s actually the smallest part of what I use. The ongoing work with KEPA has meant Palisade’s products are a very integral part of the success of this programme. Pretty much if we don’t use @RISK at least once a day, it means we’re on vacation.”

In 2012, the Kuwait Environment Public Authority (KEPA) issued new air quality regulations to categorise sources which cause, or contribute significantly, to air pollution. As part of the regulations, KEPA also created the Kuwait Implementation Plan (KIP) to ensure compliance with Kuwait Ambient Air Quality Standards (KAAQS). As part of the KIP, Kuwait was divided into Air Quality Control Zones (AQCZ) based on monitored ambient air quality.

Creating a benchmark inventory using Palisade’s risk analysis toolset

Consultants from United Nations Development Program (UNDP) assisted KEPA to create a national emissions inventory to provide a countrywide air modeling capability that would allow a source permit and monitoring programme. At the outset of the project, UNDP found there was a vast amount of organisations within the region with industrial processes that created air emissions, but that recordkeeping was unusable. In many cases, required parameters such as atmospheric pressure and flow rates were not recorded. To overcome this lack of data, UNDP turned to Palisade’s risk analysis software @RISK to capture the full range of possible outcomes as opposed to establishing one particular static number. Palisade’s risk analysis software gave UNDP the ability to justify the results of its research to KEPA by showing the range of results. This allowed KEPA to understand what data they really needed in order to ensure correct measurement of emissions.

Helping Kuwaiti businesses understand their contribution to air quality

Understanding contribution to air quality is complex. Wind conditions and production output from industrial sources can contribute to a high concentration of pollution in an area. When the weather changes, the concentration changes as a result. Brian Freeman, Team Leader for Air Regulatory Management Systems at UNDP, identified @RISK as an integral part of facilitating the success of the project. “As part of the analysis, we needed to provide input parameters into complex air dispersion models. These parameters included physical dimensions such as stack heights and diameters that were easy to identify. The more important parameters were independent variables such as release rates, temperature, and flow rates. This was where Palisade’s @RISK was particularly helpful. We were able to model the worst-case to best-case scenarios for each parameter in our models to show long-term exposure, as well as contributions from individual sources.”

Using Monte Carlo Simulation is fundamental to success

In this ongoing project, UNDP employs @RISK to correlate local industries’ impact on air quality in Kuwait. Using Monte Carlo Simulation, UNDP looks at potential ranges of impacts. The analyses can show quite conclusively, for instance, that a particular area with specific companies may not be the sole contributors to poor air quality, and are contributing significantly less than was initially considered. Two different results graphs offer a worst-case, with a particular factory contributing 4% on a bad day. But 95% of the time, based on wind patterns, that factory’s contribution was less than 2%. UNDP was also able to take years of historical monitoring data and fit the information into distribution models. The resulting graphs were output as PDFs, and allowed for year-by-year comparisons to show improvement or deterioration of pollutant concentrations. Comparative statistics were incorporated to determine overall improvement by using the mode and the 95th percentile.

Brian Freeman reports that Palisade’s software “is an essential tool that is used daily in our organisation.”

"The ongoing work with KEPA has meant Palisade’s products are a very integral part of the success of this programme. Pretty much if we don’t use @RISK at least once a day, it means we’re on vacation."

Brian Freeman
Team Leader for Air Regulatory Management Systems, United National Development Program

Favoured features and distributions

Binomial distribution – UNDP uses binomial distributions as switches – P1 representing one condition, P2 representing another condition. In many instances they see data from the same source follow one distribution over one set of conditions and follow another distribution over another set of conditions. They might not know what those input conditions are but the results are clear when shown on a histogram. Using a binomial distribution they can assign a probability that Condition P1 might happen and a probability that Condition P2 might happen. If Condition P1 happens, they use PDF 1, if P2 happens, they use PDF 2. The following example showing measured Sulfur Dioxide (SO2) shows how UNDP used the binomial distribution to create a composite PDF.

Triangle distribution – UNDP uses the triangle distribution when the underlying distribution of a data set is not known, however it has finite boundaries. It is very good for first-draft models when little is known about the system and limited data points are available.

Results graphs demonstrate significant changes – Below is a model demonstrating the overall yearly improvement or deterioration of ambient air. In this case, overall SO2 has decreased by 15%.

Freeman concludes, “I’ve used @RISK within my work for over 10 years. Nothing else matches the flexibility and ability of the software. Other products lack the interface ability completely, and whilst MCS is useful, it’s actually the smallest part of what I use. The ongoing work with KEPA has meant Palisade’s products are a very integral part of the success of this programme. Pretty much if we don’t use @RISK at least once a day, it means we’re on vacation.”

magnifierarrow-right
linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram