Estimating System Demands

From Raw Data to Base Demand

The starting point for any demand calculation is the concept of a unit demand. This is the average water consumption for a single unit, which could be a person, an area, or some other metric. To understand the terminology used in WaterGEMS, let's refer to a key Bentley resource.

As the article explains, we typically work with two main types of unit demands:

1. Per Capita Demand: Based on population (e.g., liters per person per day).
2. Area-Based Demand: Based on land use type (e.g., gallons per minute per acre of commercial property, or an equivalent population per hectare).

A client project will often involve referencing local municipal design standards to find these values. Let's look at a real-world example to see exactly how this works.

This document gives us all the components needed to calculate the Average Day Demand (ADD), which serves as our Base Demand in the model. The core formula is:

For example, using the values from the Saskatoon document for a 10-hectare low-density residential area:

• Area = 10 ha
• Density (from Table 3-1) = 42 people/ha
• Unit Consumption = 290 L/capita/day

WaterGEMS typically requires flow in units like liters per second (L/s) or gallons per minute (gpm). You must perform this conversion.

This calculated value is the "Base Demand" you would enter into the junction's Demand Collection, which we saw in the previous lesson.

Allocating Demands to Model Nodes

Once you've calculated the total demand for a service area, the next step is to distribute it among the junctions (nodes) in your WaterGEMS model. There are several ways to do this, ranging from simple to sophisticated.

Method 1: Lumped Demand

For a very small or simple area, you might assign the entire calculated demand to a single, central junction that represents that area. This is quick but less accurate, as it doesn't reflect how water is drawn throughout the area.

Method 2: Uniform Distribution

If an area is served by multiple junctions (e.g., a subdivision with 5 junctions), you can simply divide the total demand equally among them. Using our previous example with 5 junctions:

This is more realistic than the lumped method but still assumes uniform density, which may not be the case.

Method 3: Area-Based Allocation (GIS/CAD Approach)

This is the most accurate and common method used in professional practice, especially for freelance projects where clients provide spatial data. This method uses polygons to assign demands from different land use areas to the nearest junction.

The WaterGEMS tool LoadBuilder is designed for this, and we will cover it in detail in a later module. For now, it's crucial to understand the concept.

This image perfectly illustrates the idea.

As the image shows, the process is as follows:

1. Define service areas: Your client provides GIS or CAD files showing different land use zones (e.g., residential, commercial).
2. Create polygons: The software generates polygons (often Thiessen polygons) around each demand node. Each polygon represents the geographic area that is closer to its node than to any other.
3. Calculate and assign demand: The software automatically calculates the demand within each polygon based on the land use data (Area x Unit Demand Rate) and assigns that calculated demand to the corresponding node.

This method is powerful because it directly uses the spatial data clients provide and automates the tedious task of allocating thousands of individual demands.

Practical Example: A Small Commercial Lot

Let's walk through one more quick example you might face in a freelance job.

Client Request: "Analyze the impact of a new 2-hectare commercial development. Our local standard uses the Saskatoon design guide."

1. Look up parameters (from PDF LINK):

   • Land Use: Local Commercial
   • Equivalent Population: 160 p/ha
   • Unit Consumption: 290 L/capita/day

2. Calculate Total Average Day Demand:

   $$\text{ADD}=2\text{ ha}\times 160\frac{\text{people}}{\text{ha}}\times 290\frac{\text{L}}{\text{person}\cdot \text{day}}=92,800\frac{\text{L}}{\text{day}}$$

3. Convert to Model Units:

   $$\text{Base Demand}=\frac{92,800\text{ L/day}}{86,400\text{ s/day}}\approx 1.07\text{ L/s}$$

4. Allocate to Model: You would then assign this 1.07 L/s demand to the junction(s) serving this new commercial lot, using one of the allocation methods discussed above. You would also be sure to apply a commercial diurnal pattern (as discussed in the last lesson) to this specific demand.

Conclusion

In this lesson, we bridged the gap between raw client data and the actionable base demands needed for a hydraulic model. You now have a systematic process for estimating water demands, a skill that is at the heart of nearly every water distribution modeling project.

Key Takeaways:

• Base Demands are calculated from unit demands, which are typically based on population (per capita) or land use (per area).
• Real-world unit demand values are found in municipal design standards or other client-provided technical documents.
• The basic formula is: Total Demand = Area × Density × Unit Consumption Rate.
• Once calculated, total demands must be allocated to specific nodes in the model. This can be done manually for simple cases or automated using area-based methods (like the Thiessen Polygon approach) for larger, more complex systems.

You now understand the two components of demand: the base demand (what we learned today) and the demand pattern (what we learned previously). Combining these allows for a realistic simulation of water consumption.

In the next module, we will shift from concepts to application. We'll begin the process of building a model from scratch, starting with the first task in any freelance project: taking the client's raw data files (like CAD drawings and GIS shapefiles) and preparing them for use in WaterGEMS.