Step 4: Load matching
Why should you perform a load match?
Load matching is one of the single most important tasks to perform in a Conceptual study, as it assures that your digital twin works as similar to the real-life situation as possible.
To explain exactly why, let’s take a look at the data from the previous project, shown here as load duration curves of the gas usage.
The line in blue shows the real-life situation in 2019, the red line shows a non-load matched model, and the blue line shows the finished load match. As you can see, the red line is not even close to the real situation. The peak of the installation is 3 times higher than the real-life situation and for almost 6500 hours in the year, the installation will use more gas than the actual situation. (3,5 million kWh compared to 9,5 million kWh, again higher).
As you can imagine, this total gas usage compared to the real situation will also cause significant differences in the following study. Any savings that will be calculated will be done compared to this digital model, and hence, have no real correlation to the actual situation.
How do you perform a load match?
Step 1: Organize received data
First and foremost the available data should be checked to see what is available. As every project and location is different, the data will also differ, but it is still of paramount importance to try and get as accurate data as possible. There are a couple of different options:
Data showing the consumption for a full year (or several months)
Data showing monthly consumption (through payments for example)
Data showing hourly, half-hourly, or quarterly gas consumption (which is the ideal situation)
Each step shows an improvement in accuracy for your load match.
Attached here you can find a template used by Hysopt to perform the load match. In yellow you can insert the gas data supplied, while in blue you can implement the gas usage from Hysopt after a yearly simulation.
This is what it looks like for the example project:
It is clear that there is no usage in summer, and hence, there should be no activation.
Step 2: Preparing and running your Hysopt model → CTVT
By now, if you have followed the previous steps, you should have a working reference model with logical flows and temperatures and you can start setting the model. How exactly you do this is depended on the type of installation you have drawn in, but the most common way will be the usage of zones, by implementing either a relative or an absolute design load on the end unit.
Important note; DHW will not be included in this part of the explanation, you can find more information in the next step.
Relative method
If you click on a zone you see a couple of different parameters, the most important of which is “Design load”, where you can insert either an absolute or a relative load compared to the design. By implementing a relative negative percentage, Hysopt lowers the required heat of the room compared to the design, and hence, by running multiple simulations at different percentages, a value can be found which reaches the same total gas usage. For the example, the zones were found at -78,5%.
An easy method to do this is quick, is by using the sensitivity analysis tool on this design load. You could run the first one with values (-10,-30,-50,-70, and -90%). In the following Pareto analysis, you can find the total gas consumption and compare it to your total consumption.
Absolute method
Alternatively, you can fill in the absolute value. Instead of calculating percentages, you could simply fill in 2kW on a 3 kW radiator and simulate the same process. However, to save time, the relative method is more often used.
This, however, does not mean the absolute method is never used. Whenever optimisations are made, the load match will vary slightly, certainly whenever downsizing is considered. After producing your original load match, for example -78,5%, you could calculate the exact value and fill this in on the absolute value. This should ensure a close load match.
Step 3: Preparing and running your Hysopt model → DHW
Domestic Hot Water will not be controlled by the same zone model as CTVT-circuits. For this purpose, there are a couple of other blocks, known as tap profiles (both standard and eco-design). In these blocks, you can set a flow demand, and hence, by changing this demand, the DHW portion of the circuit can be load matched to the reality.
Step 4: Extracting and comparing the Hysopt monthly gas data
By opening the graphs for the boilers for “energy consumption integral” and exporting the data to Excel, you get to an Excel sheet with information on the 4 boilers, from left to right, with each showing hourly, daily, weekly and monthly usage of the boilers.
Hence, the sum can be made for monthly data to find the total usage, and compare it to the total usage given. For 0% Filled in, the following results can be found for the example
The calculated 22% shows that the system is currently significantly oversized. By using the method described in step 2, a percentage of -78,5% was found, which correlated to the following results, showing an excellent load match.
Step 5: Load matching the load duration curves
This final step is only applicable if the data given is accurate enough (hourly, half-hourly and quarterly).
By sorting the data from highest to lowest (based on hourly as this is the most accurate you can use from Hysopt) you can see the load duration curve. By exporting the hourly data and sorting the data from largest to smallest we can compare the two, and find more information on peak loads.
The results, and hence, possible changes to make to the model will change depending on the project, but in general, the important changes consist of the activation. If your entire building activates at the same time, the generated peak will be larger than what is most likely needed in reality.
Take the example above where a very clear peak in the system is found at 8h in the morning, at the activation of the entire system. If we now divide this activation over a 2-hour system (while keeping the total activation the same) we can lower this peak, as you can see below, to a more steady flow of heat throughout the day.
The extent of the changes will depend on the specific installation and information received but in general, you should end up with a near-to-100% Yearly total load match, with a similar flowing monthly usage and a correlating load duration curve. If this is the case, you have completed the load match, and you can continue with the optimisations.