Choosing a DNI

I'm a Chem Eng student currently completing an undergraduate design project.  I am to design a solar power tower. My project is different from the standard application of SPT in that I aim to use my plant to generate two steam streams; one saturated steam stream for process heat (condensate returned) and one superheated steam stream to be consumed in a chemical reaction in a theoretical acrylic acid plant. I aim to use SAM to help me with process sizing, by working out the MWthermal I require and trying to use that as an input (even if this means i have to fiddle with the MWelectircal and efficiency to get this value as an input) My question is in a few parts

-When choosing a DNI for electricity production, why is the best day often chosen as the design point, wouldn't choosing an average or worst case DNI provide a more consistent annual MWthermal?
- My aim is to produce the required flowrates of the two steam streams as consistently as possible throughout the year, is there a way I can choose my design point DNI to optimise this? (p.s. I have costed a backup natural gas plant which will need to be installed for support)
-Does this application sound stupid to you? The theoretical plant I am supplying requires electricity (but I though using SPT for this ALSO might complicate things too far) so If the idea of providing process heat through this is stupid, I could speak to my sueprvisor about instead using an SPT in the conventional manner to procude the electric power needs? 

Many, many thanks if anyone from NREL is able to help out with this long winded question

Kind regards

Harry

Hi Harry,

The choice of design-point DNI determines the relationship between the rated power cycle capacity, which is the system's nameplate capacity in MWe and the total collector area. The guidelines we provide in Help are one standard approach, but you might have reasons to use a different approach.

A better approach might be choose the design-point DNI based on the weather data (typically around 950 W/m2 for sunny areas like the U.S. Southwest and lower for cloudier areas), and adjust the total collector area using the solar multiple to try to get as constant an output as possible.

Best regards,
Paul.