In this article you're going to learn how to delete peaks on the EMS platform.
Content of this article:
This app can be used to remove peaks due to a big change in cumulated readings, such as a meter change. Learn more about Manual Readings app.
If your meter has basic cumulated data, and there is a big change in the readings which is maintained (this usually happens when utilities change meters), the readings cannot be eliminated as a single peak (just deleting one reading) since the peak would slide in time. In this case, the consumption for the period in which the reading jumps must be set to a constant value (such as 0) so that the peak does no longer appear.
For this, you should first identify the approximate period in which the peak occurs through a graph in the consumption screen:
To identify the exact moment when the peak occurs, you must follow the following steps:
- Go to the "Settings" menu and access the "Gateways" section.
Go to the "Delete Readings" tab
In this example, the peak occurs between 7:00 and 07:15 on 19/11/2018.
The readings that must be corrected are therefore the ones that correspond to the value before the peak, 19/11/2018 07:00. Taking note of this moment you can set the consumption reading to 0 for each frequency for which you need to perform analysis through the app.
Following the steps described in this article about the Manual Readings app, the corresponding gateway is configured.
Once the gateway is configured as required by the application, you can select the location and the parameter and you can insert readings with 0 value for each frequency (Quarterly: 15 minutes, Half hour: 30 minutes, Hourly: Hours, etc. ...) as seen in this caption:
Once the reading is imported, if you wait for about 5 minutes, the peak for that frequency in the consumption graph will have disappeared. In the example, the quarter-hour frequency has been used. The process must be repeated for each frequency that is causing problems.
Once you have identified the peak, the type and the time, you can delete it through the Delete readings tab. Learn more about delete readings.
In order to delete a single peak follow this steps:
1. Identify the peak and the gateway that contains the device.
2. Go to gateways -> Delete readings
3. Select the parameter and input the dates that contain the timestamp where the peak occurs and review the readings.
4. Select any wrong readings that appear and click on "Delete". The spike should disappear from your chart after a short delay. If it is not the case, try manually recalculating the device for the period there the spikes happened.
Note: This procedure works for single peaks from discrete readings.When the wrong value happens once. When you delete the reading the spike is corrected.
There are some cases where data must be restored and reinserted again in the platform to avoid the peaks. This can be do in to ways through excel or through our open API. Learn more about Importing data to the EMS platform.
Here you have cases where is necessary to import data to delete peaks:
In order to restore this kind of peaks you should follow this steps:
- Download the Active Energy readings in Excel
- Treat the data in order to avoid the peak.
- Delete the wrong readings
- Import new data on Excel using the EMS platform template or inserting the new values through the API.
You are inserting basic 15min data of Active Energy(402). For some reason, the counter one day measures more than normal and a peak appears. In this case, for example, the 6/5/2022 at 15:30.
You can see it in different places on the platform. For example:
- As data in: Settings > Gateways > Your gateway > Delete readings > Choosing the device and dates
- As discrete data in consumption screen for all frequencies smaller than 15min (hourly, daily, weekly, yearly)
- As raw readings in Advanced Analytics with frequency "b"
To actually correct basic data, all data before the peak should be uploaded again with the shift summed. As it explains here.
However, if what we want is to correct charts that show consumption, there is a quicker alternative. This one is based in the fact that, in DEXMA, discrete data overwrites basic data.
Knowing that, only one discrete value has to be inserted in the platform. One value that matches the original curve.
In the case of the example, we'll upload a discrete 15min value (40201) for the 6/5/2022 at 15:15 with value around 1kWh.
"Upload data" is usually done through inserting Excel.
Once data is uploaded, the peak has disappeared in the 15min frequency consumption charts:
For the peak to disappear also from the other frequencies, you need to recalculate the parameter. In this case, Active Energy. This is done by:
1. Go to settings.
2. Go to Parameters (left menu)
3. Click on Recalculate for the chosen parameter (in this case Active Energy)
4. Select the chosen device and dates.
It can take some minutes but the peak will disappear from all frequencies.