Very high CIBSE Solar Gains

I modelled a simple 5mx5m glass box with no heat transfer across the floor or ceiling and I created a custom glass to replicate a Pilkington glass (U-Value of 0.6 and g-factor of 0.55).  When I checked the solar gains using CIBSE Gains they were 10.8kW.  However when I calculate the gains manually using the CIBSE guide I calculate the solar gains to be 6.2kW.

Furthermore the CIBSE Gain results on Heavacomp gives the average solar gain as 410W/m2 where as in the CIBSE guide, 410W/m2 is the solar gain for a southern aspect only (E&W are 189W/m2 and north is 117W/m2).  Therefore this is leading to conclude that Heavacomp has not adjusted the solar gains for each wall aspect but rather used the max solar gains for all four walls.  If I do the same in my hand calcs, I get a total solar gain of 11.3kW so pretty close.

Also I have noticed that the CIBSE Solar Gains are double the Carrier and ASHRAE Solar Gains.

Can you think of an explanation for the large discrepancy for the two values?

Gary