Author: Emmet Ashe, student, Bachelor of Engineering in Chemical & Biopharmaceutical Engineering (Level 8), Department of Process, Energy and Transport Engineering, Cork Institute of Technology (supervisors - Noel Duffy and Dr Aisling O’ Gorman) The INTRAWaste model was designed on behalf of the Environmental Protection Agency for the major waste streams in Ireland. It is an integrated waste-management programme that combines waste streams, waste collection, treatment and disposal methods, with the objective of achieving environmental benefits by allowing the waste manager to compare new management scenarios with their present waste-management system. The model simply gives the decision maker some base mathematical evidence to help make a decision regarding the environmental impact of their municipal waste management plan. The project aim is to access the sensitivity of the consequences to variations in the INTRAWaste model and to the impact parameters within. With regard to software work, three new versions of the INTRAWaste model were programmed and validated to be used during the sensitivity assessment: 1) Updated INTRAWaste Model

  • Previously restrictive locked cells unlocked;
  • Transportation formulae error resolved;
  • Net tCO2e emission conversion fixed
  • Incorrect unit headings resolved;
  • Gasification SOx process data fixed;
  • Debatable Non- Biogenic tCO2e result now shown.
2) INTRAWaste Model With Monte Carlo Stimulator
  • When faced with uncertainty the user now has the option to uniformly generate over 110 parameter values obtained from the INTRAWaste Inventory data worksheet;
  • Monte Carlo applicable parameters and process data are highlighted yellow to alert the user of the optional feature.
3) INTRAWaste Super User Version
  • Contains the updated INTRAWaste model with Monte Carlo Stimulator;
  • Reduced from 20 to only three worksheets;
  • VBAprogrammed for faster navigation;
  • The user can now check parameter impact instantaneously by viewing the per cent change of net tCO2e once a single change is made to the scenario.
COMPARISON WITH AN ALTERNATIVE MODEL In order to ensure that the INTRAWaste model was not processing inputs into totally inaccurate and unrealistic outputs, the INTRAWaste model was compared with the Enterprises pour l'Environnement (EPE) model. The Dublin region waste management system was selected as the base scenario. [caption id="attachment_12438" align="alignright" width="873"] Fig 1. Column chart conveying the differences between the two model results (click to enlarge)[/caption] As shown from the column chart (right), the end result relies almost fully on the amount of tCO2e calculated from the landfill facilities and the amount avoided from reprocessing material.
  • The INTRAWaste model takes into account NOx emissions which is then converted to CO2e which is 54% of the overall tCO2e avoided.
  • The EPE only takes in account a straight CO2e factor. This is because the INTRAWaste model has a large inventory of data for every type of material. The EPE model is not intended to compare scenarios thus not requiring vast detail. It is simply a GHG emission reporting tool to be used on a yearly basis.
  • The EPE landfill emissions were calculated using a first- order decomposition rate equation which can be used to determine the amount of CO2e released in a year or over a range of years, while the linear INTRAWaste model does not indicate nor take into account the time horizon of its landfill emissions. In conclusion, the INTRAWaste model can only be truly accurate for scenarios compared within the models own common boundaries.
UNCERTAINTY AND SENSITIVITY ANALYSIS  [caption id="attachment_12442" align="alignright" width="962"] Fig 2. Process diagram of how the program uncertainty's are controlled (click to enlarge)[/caption] It is a given that there is no possible way of achieving factual values for waste-management models, due to the extraordinarily long period of time involved in a whole life cycle. It is best to incorporate this actual uncertainty to the programme and then analyse the results via histograms and confidence intervals. It was found that the standard deviations were very high when incorporating the landfill and recyclable relevant parameters due to the amount of numerical weight involved because of the particular scenario selected. [caption id="attachment_12445" align="alignright" width="889"] Fig 3. A line chart representing the effect of increasing the percentage of refused material prior recycling (click to enlarge)[/caption] During the sensitivity analysis of the linear model, it was found that the percentage change of contamination, and losses in paper and cardboard in the reprocessing facilities, had the most effect in this scenario. This is because the value change has an effect that propagates throughout the entire model, predominately the amount of CO2e avoided via reprocessing and the amount sent to landfilling without alternative treatment.