Interview Questions for Experience with waste forecasting and modeling

Waste forecasting models come in various forms, each with its strengths and weaknesses. I’m experienced with several, including:

• Time Series Models: These models analyze historical waste generation data to identify patterns and trends, predicting future amounts. Examples include ARIMA (Autoregressive Integrated Moving Average) and exponential smoothing models. They are effective for capturing seasonality and trends but may not adequately account for external factors.
• Regression Models: These models explore the relationship between waste generation and various influencing factors. Multiple linear regression is common, where we might predict waste based on population, economic activity, or waste diversion rates. They’re useful for understanding causal relationships but require strong data on predictors.
• Machine Learning Models: Techniques like support vector machines (SVM), random forests, and neural networks offer powerful predictive capabilities, especially with complex datasets. They can handle non-linear relationships and large amounts of data, but require significant computational resources and careful model selection.
• Agent-Based Models (ABM): These are particularly useful for simulating complex systems where many factors interact. In waste management, an ABM could model the behavior of individual households or businesses to predict aggregate waste generation. They are computationally intensive but capture emergent behavior not readily apparent in simpler models.

The choice of model depends on factors like data availability, computational resources, and the desired level of detail in the forecast.

My experience with statistical modeling in waste forecasting is extensive. I regularly use:

• ARIMA modeling: I’ve used ARIMA to forecast monthly municipal solid waste (MSW) generation for a large city, successfully incorporating seasonal fluctuations (higher waste during holidays) and long-term trends (growth due to population increase). The model’s parameters were tuned using techniques like AIC (Akaike Information Criterion) to optimize predictive accuracy.
• Regression analysis: In another project, we used multiple linear regression to model per capita waste generation. We included independent variables such as income levels, household size, and recycling participation rates. This provided valuable insights into the factors driving waste generation and allowed us to develop targeted waste reduction strategies.
• Time series decomposition: Before applying any advanced model, I always decompose the time series data to isolate trend, seasonality, and residuals (random variation). This step is crucial for accurate forecasting, ensuring that each component is properly modeled.

Beyond these, I have experience using techniques like exponential smoothing, particularly for situations with limited data or rapidly changing trends. The selection of statistical method always depends on the nature of the data and the research questions.

Uncertainty and variability in waste generation data are inevitable. I address them using several approaches:

• Robust Statistical Methods: I often utilize robust regression techniques, less sensitive to outliers, which frequently occur in waste data due to factors like data entry errors or unusual events (e.g., a major storm).
• Monte Carlo Simulation: This powerful method allows us to incorporate uncertainty in model parameters by running many simulations with randomly sampled parameter values. This generates a distribution of possible future waste generation scenarios, providing a range of predictions rather than a single point estimate.
• Bayesian methods: These offer a more formal approach to handling uncertainty by combining prior knowledge with data to generate posterior probability distributions for model parameters. This allows for more informed decision-making under uncertainty.
• Error Analysis: Thorough analysis of residuals (the differences between actual and predicted values) helps identify patterns indicating model shortcomings or unforeseen factors influencing waste generation. This allows for model refinement or consideration of additional variables.

By combining these methods, we can quantify the uncertainty in our forecasts and provide more reliable and useful information for decision-makers.

Many factors influence waste generation. In my models, I consider:

• Population size and demographics: Population density, age distribution, and household size significantly impact waste generation.
• Economic activity: Economic growth and consumer spending patterns influence the amount of packaging and disposable goods consumed.
• Seasonal variations: Waste generation often increases during holidays and warmer months due to increased outdoor activities and tourism.
• Waste diversion programs: Recycling and composting rates directly reduce the amount of waste sent to landfills.
• Waste management policies: Changes in regulations, fees, or collection services can impact waste generation.
• Special events: Major events like festivals or sporting events can temporarily increase waste generation.
• Weather patterns: Extreme weather can disrupt collection and affect the composition of waste.

The specific factors included in a model depend on the context and the availability of data. For example, a model for a rural area might not need to account for as many factors as one for a large metropolitan city.

Validation and verification are crucial to ensure the accuracy and reliability of waste forecasting models. I employ several techniques:

• Goodness-of-fit measures: Statistical metrics like R-squared, RMSE (Root Mean Squared Error), and MAE (Mean Absolute Error) assess how well the model fits the historical data. Lower values indicate better accuracy.
• Backtesting: The model is used to predict past data points that were not included in the model’s training. Comparing predicted to actual values helps gauge its predictive performance.
• Holdout samples: A portion of the available data is withheld from model training and used for independent validation, providing a more unbiased assessment of accuracy.
• Cross-validation: Multiple training and validation sets are created, allowing for a more robust evaluation of model performance and helps identify potential overfitting.
• Sensitivity analysis: Varying model parameters to understand their impact on predictions helps identify critical factors and uncertainties.

These methods allow for a comprehensive assessment of model performance and aid in identifying areas for improvement or refinement. The goal is not perfect prediction, but rather a model providing useful insights and sufficiently accurate projections to inform decision-making.

My experience encompasses a wide range of data sources for waste forecasting, each offering unique perspectives and challenges:

• Municipal records: These are primary sources, providing detailed information on waste generation, collection routes, and diversion rates. They’re usually reliable but can be incomplete or inconsistently recorded.
• Surveys: Household surveys provide valuable insights into waste generation behaviors, allowing for deeper understanding of the factors driving waste patterns. They are more expensive and time-consuming than other sources but offer critical context.
• Commercial waste data: Data from businesses provide additional insights, particularly on industrial waste streams, which differ significantly from residential waste.
• Geographic Information Systems (GIS): GIS data provides spatial context, revealing how waste generation varies across neighborhoods or regions, which can help target waste reduction efforts.
• Waste composition studies: Analyzing the composition of waste helps understand the sources of waste and allows for better targeted interventions.
• Census data: Population statistics, household income, and other demographic factors are essential for developing accurate waste generation models.

Data integration from multiple sources is crucial. Often, data cleaning and harmonization are necessary to create a consistent and useful dataset for modeling.

Seasonality and trends are critical aspects of waste forecasting. I incorporate them in several ways:

• Seasonal dummy variables: In regression models, I include dummy variables representing different seasons or months to capture seasonal fluctuations. This allows the model to account for predictable variations in waste generation throughout the year.
• Time series decomposition: As mentioned earlier, decomposing the time series into trend, seasonal, and residual components allows for a structured modeling of each component. The seasonal component can be directly included as a feature in the model, while the trend can be modeled using ARIMA or other time series techniques.
• Fourier series: For capturing more complex seasonality or cyclical patterns, I sometimes use Fourier series to represent seasonal variations as a sum of sine and cosine functions.
• Moving averages: Moving averages can smooth out short-term fluctuations, revealing underlying trends in waste generation more clearly.

The specific method used depends on the complexity of the seasonal patterns and the type of model employed. For example, a simple seasonal dummy variable approach might suffice for a relatively stable system, while more sophisticated techniques might be necessary for a system with complex cyclical patterns.

My proficiency in waste forecasting and modeling spans several software and tools. I’m highly experienced with statistical packages like R and Python, leveraging libraries such as statsmodels, scikit-learn, and specialized packages for time series analysis. These allow me to perform complex statistical analyses, build predictive models, and visualize results effectively. I also have extensive experience with Geographic Information Systems (GIS) software, specifically ArcGIS, which is crucial for spatially mapping waste generation data and identifying trends. Furthermore, I’m comfortable using simulation software, such as AnyLogic, to model complex waste management systems and test different intervention strategies. Finally, I’m proficient with data visualization tools like Tableau and Power BI to communicate findings clearly and concisely to both technical and non-technical audiences.

In a recent project for a large metropolitan area, we used waste forecasting to optimize their waste collection routes and infrastructure planning. The city was facing increasing waste generation due to population growth and changing consumption patterns. We first collected historical waste data from different collection zones, incorporating factors such as seasonality, holidays, and local events. We then employed time series analysis in R to identify trends and predict future waste volumes. Using machine learning algorithms, we built a predictive model that incorporated socio-economic indicators, population density data from the GIS, and recycling rates to project waste generation for the next 10 years under different scenarios. This allowed the city to make data-driven decisions regarding the investment in new collection vehicles, the expansion of landfill capacity, and the optimization of their existing waste collection network, ultimately resulting in significant cost savings and improved operational efficiency.

Communicating complex technical information to non-technical audiences requires a clear and concise approach. I avoid using technical jargon whenever possible, instead opting for plain language and relatable analogies. For instance, instead of saying ‘we employed an ARIMA model for time series forecasting,’ I might say, ‘we used a sophisticated method to predict future waste based on past trends.’ I utilize visualizations extensively—charts, graphs, and maps are far more effective than complex equations. I also tailor my communication style to the audience’s level of understanding, providing simpler explanations for less technical individuals and more detailed information for those with some background in the field. Finally, I always make sure the key findings and their implications are clearly summarized, focusing on the practical consequences of the predictions.

Current waste forecasting models have limitations. One major challenge is accurately predicting the impact of policy changes and technological innovations on waste generation and composition. For example, the introduction of a new recycling program or a ban on single-use plastics can significantly alter waste streams, and these changes are often difficult to anticipate with complete accuracy. Another limitation lies in the availability and quality of data. Incomplete or inconsistent data can lead to inaccurate forecasts. Furthermore, many models assume linear trends, while waste generation is often influenced by non-linear factors. To improve these models, we need to incorporate more granular data, including real-time data from smart bins and improved data collection methodologies. The integration of machine learning techniques and agent-based modeling could also enhance predictive accuracy by capturing complex interactions and non-linear relationships within the waste management system. Finally, greater emphasis should be placed on model validation and uncertainty quantification.

Scenario planning is crucial for robust waste management planning. It involves creating different plausible future scenarios based on various factors, such as population growth, economic development, technological advancements, and policy changes. For example, we might develop three scenarios: a ‘business-as-usual’ scenario, a ‘high-recycling’ scenario, and a ‘significant economic downturn’ scenario. Each scenario would have its own set of assumptions regarding waste generation, recycling rates, and landfill capacity. By modeling waste generation under each scenario, we can assess the robustness of different waste management strategies and identify those that are most resilient to uncertainty. This allows decision-makers to develop flexible and adaptable plans that can accommodate a range of potential futures.

Recycling rates and diversion strategies are fundamental to accurate waste forecasting. We incorporate them by disaggregating waste streams into recyclable, compostable, and residual fractions. Historical data on recycling rates is used to estimate future recycling behavior. However, we also account for potential changes in recycling rates based on policy interventions, public awareness campaigns, or technological advancements in recycling technologies. Diversion strategies, such as composting programs and source separation initiatives, are modeled by adjusting the proportions of waste going to different disposal pathways. For example, if a new composting program is implemented, we adjust the model to account for the expected increase in composted waste and the corresponding decrease in landfill waste. This integrated approach allows us to project the amount of waste going to different disposal options and assess the effectiveness of various diversion strategies.

Life-cycle assessment (LCA) is a critical tool in evaluating the environmental impacts of waste management strategies. It’s a holistic approach that assesses the environmental burden associated with a product or process from its cradle to grave, including raw material extraction, manufacturing, transportation, use, and disposal. In waste management, LCA helps evaluate the environmental footprint of different disposal methods, such as landfilling, incineration, and recycling. For example, an LCA might compare the greenhouse gas emissions associated with landfilling organic waste versus composting it. By quantifying the environmental impacts, we can make informed decisions that minimize the overall environmental burden of waste management systems. The results of an LCA can be incorporated into waste forecasting models to help predict the environmental consequences of different future scenarios and to guide the selection of environmentally sustainable waste management practices.

Data quality is paramount in waste forecasting. Think of it like baking a cake – if your ingredients are bad, the cake will be bad. Inaccurate or incomplete data leads to unreliable predictions. We ensure data quality through a multi-pronged approach:

• Data Source Verification: We meticulously check the source of our data, ensuring it’s from reliable and reputable organizations. This includes validating collection records, verifying weighing scales’ calibration, and checking for inconsistencies across different data sets.
• Data Cleaning and Preprocessing: This involves identifying and handling missing values, outliers, and inconsistencies. We might use imputation techniques to fill in missing data based on statistical models, or remove outliers only if they are clearly erroneous and not representative of actual waste generation patterns.
• Data Validation: We use cross-validation techniques to check the consistency of our data. For instance, we compare waste generation data with population density, industrial activity levels, and seasonal trends to identify anomalies and inconsistencies.
• Regular Audits: We conduct periodic audits to identify any potential issues with data collection or processing. This might involve site visits to waste transfer stations, review of collection routes and schedules, and interviews with waste management personnel.

For example, in a project involving municipal solid waste, we identified a discrepancy in data from a specific collection route. After investigation, we found a faulty weighing scale at the transfer station, leading to underestimated waste volumes. We corrected the data and implemented a preventative maintenance schedule for the scales.

Geographic Information Systems (GIS) are indispensable in waste management modeling. Imagine trying to plan a city’s waste collection routes without a map – it would be chaos! GIS allows us to visualize and analyze spatial data, making it far more effective than traditional methods.

My experience includes using GIS to:

• Map waste generation hotspots: Identifying areas with higher waste generation based on population density, commercial activity, and other relevant factors. This helps optimize collection routes and resource allocation.
• Design efficient collection routes: GIS helps determine the shortest and most efficient routes for garbage trucks, minimizing travel time and fuel consumption. We often use algorithms such as vehicle routing problems (VRP) solvers integrated within GIS platforms.
• Analyze proximity to waste facilities: Evaluating the location of landfills, recycling centers, and transfer stations in relation to residential and commercial areas. This assists in planning new facilities or upgrading existing ones.
• Spatial forecasting: Predicting future waste generation patterns based on projected population growth, land use changes, and other spatial variables. This involves using spatial regression models, often combined with time-series data.

In one project, we used GIS to analyze the distribution of illegal dumping sites in a municipality. By mapping these locations and analyzing their proximity to residential areas and infrastructure deficiencies, we were able to prioritize cleanup efforts and implement targeted prevention strategies.

Optimization techniques are crucial for efficient waste management planning. Think of it like solving a complex puzzle – we need the best solution to minimize costs and maximize resource utilization. I have extensive experience applying various optimization techniques:

• Linear Programming (LP): Used to optimize waste collection routes, minimizing travel distance and costs while meeting service demands. For example, we might use LP to determine the optimal number of trucks and their routes to collect waste from various locations within a specified time window.
• Mixed Integer Programming (MIP): Similar to LP, but allows for integer variables, which is useful when dealing with discrete decisions like assigning trucks to specific routes or deciding whether to open a new landfill.
• Simulated Annealing and Genetic Algorithms: Employed for complex, non-linear optimization problems, such as optimizing the location of waste facilities or developing a network of transfer stations.
• Heuristic Algorithms: Used when the problem is too complex for exact solutions, providing near-optimal solutions in a reasonable amount of time. For example, we might use heuristic algorithms to solve large-scale vehicle routing problems.

In a recent project, we used MIP to optimize the location of new recycling facilities, considering factors such as transportation costs, population density, and environmental impact. The model identified the optimal locations that minimized overall costs and maximized recycling rates.

Data gaps are inevitable in waste management. They’re like missing pieces in a jigsaw puzzle. To address them, we employ several strategies:

• Data Imputation: We use statistical methods such as mean, median, or mode imputation to fill in missing data based on available data. More sophisticated methods like regression imputation or multiple imputation might be used to address more complex patterns of missing data.
• Data Triangulation: We compare available data with other sources to fill in gaps or to verify existing data. For instance, we might cross-reference waste generation data with population census data or commercial activity statistics.
• Proxies and Estimation: If data is completely unavailable, we can use proxies or estimation methods based on similar regions or demographics. This method should be used cautiously, and the uncertainty associated with these estimations should be clearly stated.
• Data Collection Initiatives: If gaps persist, we might suggest targeted data collection efforts, such as waste audits or surveys, to fill in the missing information. This is a more costly but potentially more accurate approach.

For instance, when dealing with missing data on a specific type of industrial waste, we used regression analysis to estimate the missing values based on the production volumes of the relevant industries and similar regions.

Outliers and anomalies in waste data can significantly skew our forecasts. Imagine a single unusually high waste generation day – it could make the whole year look off! We handle them by:

• Investigation: We first investigate the cause of the outlier. Was there a special event (e.g., a holiday, a major construction project), or is there a data error?
• Data Cleaning: If the outlier is caused by a data error (e.g., a recording mistake), we correct it. If the outlier is legitimate but not representative of typical waste generation patterns, we might choose to exclude it or weight it down in the analysis, depending on the modeling approach.
• Robust Statistical Methods: We often use robust statistical methods, which are less sensitive to outliers than traditional methods. For example, robust regression techniques provide more stable estimates even when outliers are present.
• Data Transformation: Transforming the data (e.g., using logarithmic transformation) can sometimes help reduce the influence of outliers.

In one instance, an unusually high waste volume was found to be caused by a temporary surge in construction activity related to a major infrastructure project. Rather than discarding the data point, we incorporated it into our model using a time-dependent variable that accounted for this temporary increase in construction activity.

Waste forecasting and modeling present many challenges:

• Data Availability and Quality: Consistent, accurate, and comprehensive data is often lacking, especially in developing countries or for specific waste streams.
• Data Variability: Waste generation patterns are influenced by numerous factors (seasonal variations, economic conditions, population changes, policies, etc.), making accurate predictions challenging.
• Unpredictable Events: Unexpected events (natural disasters, pandemics, economic downturns) can significantly impact waste generation and disrupt forecasts.
• Model Complexity: Developing accurate and reliable models requires expertise in statistics, optimization, and GIS, and the models themselves can be quite complex.
• Computational Limitations: Analyzing large datasets and running complex simulations can require significant computing resources.

Addressing these challenges requires a holistic approach that combines rigorous data collection, advanced analytical techniques, and robust model validation. Furthermore, effective communication and collaboration with stakeholders are crucial for successful implementation and adaptation.

Measuring the success of a waste forecasting model depends on its intended purpose. We typically use a combination of metrics:

• Accuracy Metrics: These measure how closely the model’s predictions match the actual waste generation. Common metrics include Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R-squared.
• Coverage: We assess how well the model covers the range of waste generation variability, including both typical and extreme values.
• Forecast Horizon: We need to determine how far into the future the model can accurately predict waste generation.
• Practical Application: We evaluate the model’s usefulness in informing decision-making processes. For example, does the model help improve collection efficiency, resource allocation, or facility planning?
• Cost-Effectiveness: We consider the costs associated with data collection, model development, and implementation, relative to the benefits gained from improved waste management.

Ultimately, a successful model is one that provides reliable predictions that lead to tangible improvements in waste management practices and reduce environmental impact. It’s less about achieving perfect accuracy and more about supporting effective decision-making.

Understanding waste streams is fundamental to accurate forecasting. Waste streams are categorized based on their source and composition. Think of it like sorting your recycling – different materials go in different bins.

• Municipal Solid Waste (MSW): This is the everyday trash from households and businesses – food scraps, packaging, yard waste, etc. Its composition varies greatly by region and season.
• Commercial and Industrial Waste (C&I): This includes waste generated by businesses, ranging from office paper to construction debris. This stream can be quite diverse depending on the industry.
• Construction and Demolition Waste (C&D): This is generated from building and demolition projects and includes materials like concrete, wood, and drywall. Forecasting this requires detailed project information.
• Hazardous Waste: This includes materials that pose a threat to human health or the environment. Proper handling and forecasting require specialized knowledge and regulations. Examples include batteries, chemicals, and medical waste.
• Special Waste: This category often includes waste that doesn’t fit neatly into other categories, such as bulky waste (furniture, appliances), electronic waste (e-waste), or green waste (yard trimmings).

Each stream has unique characteristics impacting forecasting. For instance, MSW is highly seasonal, influenced by holidays and weather. C&I waste is more predictable but varies widely based on economic activity.

Population growth and economic factors significantly influence waste generation. Imagine a city experiencing a population boom – more people mean more waste!

We incorporate these factors using statistical models. For population growth, we use demographic projections from reliable sources like census data. These projections help us estimate the future number of waste generators.

Economic factors are incorporated through indicators like per capita income and economic activity indices. A growing economy often correlates with increased consumption and thus more waste. We might use regression analysis to establish the relationship between economic indicators and historical waste generation data. For example, we might find a strong correlation between retail sales and packaging waste.

These factors aren’t simply added; they’re integrated into sophisticated models that consider their interactions. For instance, a growing population with higher per capita income may generate more waste per capita than a slower-growing population with lower income.

Different waste collection systems significantly impact forecasting accuracy. The efficiency and frequency of collection influence the volume of waste accumulating at any given time.

• Curbside Collection: Common in residential areas, this system’s efficiency influences how much waste is stored before collection. Factors like collection frequency, bin sizes, and participation rates affect forecasting.
• Automated Collection: Using specialized trucks, this system is efficient but may require adjustments in forecasting due to changes in bin capacity.
• Drop-off Centers: These allow residents to dispose of waste at designated locations. Forecasting requires understanding user behavior and the capacity of the drop-off centers.
• Underground Waste Systems: These systems reduce odors and visual clutter but require a different forecasting approach, as waste is compacted underground, and collection frequency may be lower.

For example, if a city switches from weekly to bi-weekly curbside collection, we’d expect to see an increase in waste accumulated between collection days, affecting the overall volume recorded.

Our models account for these variations by incorporating data on collection methods, frequencies, and efficiency rates for accurate forecasting.

Staying updated in this field requires a multi-pronged approach. It’s a dynamic field with constant advancements in data analytics, modeling techniques, and waste management technologies.

• Professional Associations: Active participation in organizations like the SWANA (Solid Waste Association of North America) provides access to conferences, publications, and networking opportunities.
• Academic Journals and Publications: Staying abreast of peer-reviewed research in journals focusing on environmental science and engineering is crucial.
• Industry Conferences and Workshops: Attending conferences allows for knowledge sharing and exposure to the latest software and modeling techniques.
• Online Resources and Databases: Utilizing reputable online databases and resources containing relevant data and information is vital.
• Collaboration with Experts: Engaging with other professionals in the field for collaboration and knowledge exchange.

By combining these strategies, I ensure I’m always up-to-date on the most effective and reliable forecasting methods.

Ethical considerations are paramount in waste forecasting. Transparency, data integrity, and the responsible use of predictions are key.

• Data Privacy: Ensuring the privacy of individuals and businesses whose data is used in the models is crucial.
• Data Accuracy and Transparency: Using accurate, reliable data and clearly communicating the limitations of the models is essential to avoid misleading conclusions.
• Bias and Equity: Models should be carefully evaluated for potential biases that might unfairly impact specific communities or groups.
• Environmental Justice: Forecasting should contribute to fair and equitable waste management practices. For instance, it should not disproportionately burden vulnerable communities with waste facilities.
• Model Validation and Uncertainty: It’s crucial to validate the models and communicate the inherent uncertainties associated with predictions.

For example, if a model predicts a disproportionate amount of waste in a low-income neighborhood, we must investigate if this is due to actual waste generation or other factors, like inadequate collection services. Addressing such issues ensures fairness and promotes environmental justice.

In a previous project involving a large-scale landfill expansion, our waste forecasting model predicted a faster-than-expected landfill capacity depletion. This conflicted with the city’s initial projections and presented a difficult decision.

The city council, initially hesitant to accept our findings, was concerned about the economic implications of a premature landfill closure. However, ignoring our data risked a major environmental and public health crisis. We presented our data transparently, emphasizing the statistical rigor and potential consequences of inaction.

After thorough analysis and several meetings, we collaborated with the city to implement an interim waste reduction strategy while expediting the landfill expansion process. This included public awareness campaigns focused on recycling and waste reduction. It was a challenging but ultimately successful situation that demonstrated the importance of acting on accurate forecasting data, even when faced with difficult decisions.

Forecasting waste for a new development requires a multi-step approach. It’s not just about extrapolating existing data; we need to consider the specific nature of the project.

1. Project Details: Gathering thorough information on the project’s scope, type (residential, commercial, industrial), density, and anticipated occupancy rates is paramount.
2. Waste Characterization Studies: Conducting detailed waste characterization studies for similar existing developments can provide valuable insights into the types and amounts of waste likely to be generated.
3. Per Capita Waste Generation Rates: Establishing baseline per capita waste generation rates from comparable developments is crucial. These rates can be adjusted based on project specifics, such as the presence of recycling programs or waste reduction strategies.
4. Modeling and Simulation: Using statistical models, potentially incorporating population projections and economic factors, we can develop scenarios for waste generation over the project’s lifecycle.
5. Sensitivity Analysis: A sensitivity analysis will assess how changes in key assumptions (e.g., occupancy rates, recycling participation) affect the forecast, providing a range of possible outcomes.
6. Regular Monitoring and Updates: Following construction and occupancy, monitoring the actual waste generation provides valuable data for model calibration and refinement.

This systematic approach ensures that our forecast is as accurate and reliable as possible, informing responsible waste management planning for the new development.