Water balance error due to irrigation

Hi,
For my coupled MIKE SHE/MIKE 11 model, i am getting water balance error of about 3% of total rainfall (in WB tool). This error is caused due to inclusion of irrigated areas in the model, as i found after running the model with and without irrigation. I have scattered irrigated areas (about 20 chunks) in the model, where each chunk has 2-3 crops. The crops and irrigation scheduling are same in each of the chunk but these chunks are irrigated by corresponding wells sources in their area. As i noticed, the error is especially high if i include irrigated paddy in the model, which requires relatively higher irrigation. I tried changing the UZ and other components time steps and also change the demand type, but it didn't work.
The error is positive and, as i understand, it implies that model is generating water. This error may mask the effects of some of the scenarios i am simulating and I was wondering if there is a way i can reduce this error? Or what is the reason for this error?
Thank you,

You could try to understand better the source of the error by producing Map Outputs from the Water Balance Calculation Tool. The water balance type "Map Output: Total error" with type "Accumulated" would give you an idea of the area with biggest errors. If you have several Irrigation command areas, you would know where to take a closer look. Other types of Map Outputs can tell you if the error is mostly produced in the OL, UZ or SZ component, and in the latest case, in what SZ layer.

If you are using single or shallow well as your source type, you may want to avoid that your SZ cells become dry.  So, the "Max. rate" can not be to high, the screening interval need to include aquifers (and not be inside a confining unit), and you can set the Max depth to water enough above the bottom.

I hope this helps.