Diurnal Demand Patterns and Customer Types

The Problem with Average Demand

Imagine a client asks you to model their system. You might know the total amount of water used in one day. Dividing this total volume by 24 hours gives you the average daily demand.

However, designing a system based on this average is a critical error. Why? Because no one uses an "average" amount of water every hour of the day.

To understand the different types of demand and their importance, please watch this short video.

As the video explains, usage fluctuates significantly. This 24-hour cycle of demand is called a diurnal pattern or diurnal curve. A model that only uses average demand might show that the system works perfectly, while in reality, residents could be experiencing low pressure every morning during the "peak hour."

Understanding the Diurnal Curve

The diurnal pattern is typically represented by a series of multipliers applied to a base demand (which is often the average demand). The multiplier is 1.0 at the time of average demand, greater than 1.0 during peak periods, and less than 1.0 during low-use periods.

This graph shows a classic diurnal curve for a residential area.

Notice the two distinct peaks:

1. A morning peak (around 7-8 AM) as people get ready for work and school.
2. An evening peak (around 6-8 PM) as people return home, cook dinner, do laundry, etc.
   Demand is lowest in the middle of the night.

Now, let's get the official WaterGEMS terminology for this concept. The following resource from Bentley explains it perfectly.

The key takeaway is the simple but powerful formula WaterGEMS uses during an Extended Period Simulation (EPS):

One Size Does Not Fit All: Patterns by Customer Type

A freelance project will rarely involve a system serving only one type of customer. A typical water network supplies residential neighborhoods, commercial districts, and industrial parks. Each has a unique pattern of water use.

• Residential: As we saw in the diurnal curve graph, this pattern is characterized by morning and evening peaks.
• Commercial (Offices, Retail): Demand is highest during business hours (e.g., 9 AM to 5 PM) and very low during evenings, nights, and weekends.
• Industrial: This can vary widely. A factory running 24/7 might have a nearly flat, constant demand. A plant that performs a specific water-intensive process once per day will have a very sharp, defined peak.

Applying the correct pattern to the corresponding part of your model is crucial for accuracy. If you assign a residential pattern to a commercial zone, your model will incorrectly predict high water use in the evening when the businesses are actually closed.

How to Create and Apply Demand Patterns in WaterGEMS

Now for the practical application. We'll use a video to walk through the process of creating and assigning different patterns for residential and commercial customers within WaterGEMS.

Let's recap the workflow shown in the video for creating a new pattern:

1. Go to Components > Patterns. This opens the Patterns manager dialog box.
2. Highlight the Hydraulic folder on the left and click the New icon.
3. Rename the pattern to something descriptive (e.g., "Residential - Single Family" or "Commercial - Office Park").
4. Ensure the Pattern Format is set to Stepwise. This means the multiplier remains constant until the next time step.
5. Enter your Time from Start and Multiplier pairs in the table. The 24-hour cycle must start and end with the same multiplier.

Here is an image of the dialog box you'll be working with. Notice how it contains a list of patterns, a table for data entry, and a graph for visualization.

Once created, you can assign the pattern to a junction by selecting it from the Pattern (Demand) dropdown menu in the junction's Demand Collection properties.

For more complex models, the video also briefly mentioned the Demand Control Center (under the Components tab), which is a powerful tool for assigning base demands and patterns to many junctions at once in a tabular format.

Conclusion

In this lesson, we moved from the static world of steady-state analysis to the dynamic world of time-varying demands. You learned that modeling how water usage changes over a 24-hour period is fundamental to understanding true system performance.

Key Takeaways:

• Diurnal demand patterns model the daily fluctuation of water use using time-based multipliers applied to a base demand.
• Relying on an average daily demand is insufficient as it misses the critical peak hour demands that stress the system.
• Different customer types (residential, commercial, industrial) have distinct demand patterns that must be applied to the correct zones in your model for accurate results.
• In WaterGEMS, you can create, manage, and assign these patterns through the Patterns manager and a junction's Demand Collection properties.

You now understand how to define the base demand and the pattern that makes it vary over time. But where does that base demand number come from in the first place? A client won't just give you a list of demands for every junction. They will give you population data, land use maps, or billing records.

In our next lesson, we will focus on estimating average system demands using population data, land use information, or other client-provided metrics. This will complete the picture of how to define and allocate demands in your hydraulic model.