Hi. I am fairly new to modelling. Therefore, having many question related to the model building as well as understanding proper ways to carry out modelling.
Here is what I have and what I did.
1. Customer meter with billing data, laterals, taps. - All the data is imported into the model and the model is built.
2. Two inlets having a known flow values in 5 meter interval, along with the pressure - These inlets are modelled as Reservoirs. The Pressure values are used to calculate the HGL and then Reservoir HGL Pattern is created and assigned to each of the inlet reservoirs respectively.
3. Pressure Loggers inside the DMAs.
For creating the demand pattern, I used the inlet flow data and created a multiplier for each of the inlet. Have created the hydraulic pattern of the same in WaterGEMS
I assigned one of this multipliers to all the customer demand nodes.
Also note that, I have calculated the Losses by [Water supplied - (Water Billed per month/30 days) ].
Then I calculated the losses per customer meter and added that value to all the meters.
Question:
1. Is the method of calculation of the demand pattern correct? If not, please suggest a correct way to arrive on it.
2. In case of multiple inlets, how do I know which inlet is supplying to which part of the network, and thereby how do I know which hydraulic pattern to be assigned to which customer?
Thanks for bearing me.
It sounds like you are modeling 1 area which has 2 inlets (so no closed valves inside the area to split it). The water consumption of the entire area is equal to Inlet 1 Flow + Inlet 2 Flow. This gives you the total water consumption (including leakage) in the area. Substract the leakage from this amount and assign it equaly over the customer meters and construct a demand pattern based on the actual water constumption (Total water consumption - leakage) and apply this to all customer meters. Watergems will then calculate the flow supplied by each inlet, if your roughness values and demand allocation is correct (hence when your model is properly calibrated), these values should be similar to your measured inlet flows.
Thanks for the reply Joeri.
I am having the customer meter billing data. I went the other way. I did :- Summation of water supplied from each inlet - summation of water billed at all consumer ends = Leakage.
And I applied this leakage to all the customers.
As you can see from the snaps, the demand pattern which I have considered is the multiplier that I got by dividing the volume at unit time with the total volume per day. This is applied to all customer meters in the model.
But, only one pattern is possible to be applied to all the customer meters. What if there are more than 2 inlets. How to apply the pattern in that case?
Thanks.
If you distribute the calculated demand at all customer meters you have the total demand (after subtracting the leakage). Run your model with both the reservoirs open and flowing freely and then check the flow contributions from both the individual reservoirs. As Joeri suggested, if your roughness coefficients and demand allocation is correct, WaterGEMS will calculate the flow supplied from each reservoir.
Once you have the calculated values, compare them with your observed values. If you have three reservoirs (more than two), WaterGEMS will calculate how much water each one is supplying based on your inputs.
Regards,
Yashodhan Joshi
Hi Yashodhan.
I am unable to follow.
I do not have a leakage data.
I have a customer billing meter data for each consumer separately ( i.e consumption)
I have a Total Supplied Volume data (i.e. inlet 1 + inlet 2) with 5 min time interval. So this is including all losses
Therefore, Total Supplied Volume - Billing meter data for all consumers = Losses.
This Losses are now equally distributed to all the consumers.
Whereas, as you are suggesting, that I should substract leakage from total volume supplied, and then remainder i distribute to all the customers... I am unable to understand, why I should distribute the demand equally to all customers, when I am having the real values.
Would like to send the model and calculations for clarity. .
Hi Chetan.
From what I've been able to get from your questions and the replies, it seems to me that you have set up your model, customer meter demands, and loss allocation correctly. I have a small municipal system in which I've modeled the demands in almost the exact same way as you describe and it works great.
As you are aware, water system losses are fairly easy to quantify on a system-wide basis, but almost impossible to quantify on an individual meter basis, so the best solution I feel is to do what you did - equally distribute losses to the meters.
One issue you mentioned but has not been addressed (that I've seen) is "only one pattern is possible to be applied to all the customer meters." From my experience this is true, however, within that pattern you can modify the hourly, daily and monthly multipliers. Since your data looks like it is fairly granular, you can calculate and apply hourly multipliers (if you haven't already done that). Weekly and monthly multipliers can also be calculated and entered. From my limited understanding these additional multipliers will be applied to the demand based on your EPS run date and time parameters.
Perhaps I'm missing something here (which is entirely possible), but your approach makes sense to me. Good luck!