Deviation of the circular area of a buffer

[milla]

I have created buffers that I now want to intersect with a layer. My problem is that the area inside the buffer doesn't match the theoretical area for a circle of that radius. I have the same deviation for every buffer. All coordinate systems are the same. The layer I'm using is definitely gapless (I checked that). The buffers that I use for intersecting have the correct area.

I need the correct area in the buffer to be able to continue with my task. Unfortunately I'm clueless.. I would be grateful for ideas and help.

[DavidPike]

What coordinate system is it in?  Could you provide the observed difference and screenshots also please?

[DanPatterson]

If the output was to a shapefile, then you have an "ngon", an n-sided replica of a circle since shapefiles don't support true curves (eg a circle) like featureclasses in a geodatabase.

If your coordinate system, is web mercator, then don't use it, use a better planar projected coordinate system.

You should be using a geodesic buffer also

How Buffer (Analysis) works—ArcGIS Pro | Documentation

Buffer (Analysis)—ArcGIS Pro | Documentation

To see if your buffer is a true curve, use 

Calculate Geometry Attributes (Data Management)—ArcGIS Pro | Documentation

and calculate Area_geodesic and CURVE_Count to see if the buffers are truly circles rather than n-gons