Details details. The devil is in the details as always.
First a little more about my heating system. I have an above the subfloor type radiant system (Viega Climate Panel) that is about 540 sq ft., 6 steel panel rads on the main floor and one zone of BB for the second floor which has a Taco 571 zone valve controlled by an ancient T87 thermostat. It is the only on/off load in the heating system. There is no dedicated heat in the basement.
As originally installed with the gas boiler, the system water temp is determined by outdoor air temp. There is a separate mixing station that runs the radiant floor temp also based on what is going on outside but it is set to a different heating curve than the main water temp.
There is only one circ for the "high temp" loads which are the BB and panel rads. The mixing station has its own circ for the radiant floor. Both heating circuits are constant circulation and controlled only by outdoor temp for shutdown.
The boiler control shuts down both heating circuits when there is a call for domestic hot water from the 60 gallon indirect and a dedicated pump runs that load.
During a typical heating season I'll see main system temps run from 120 up to about 170*. The radiant floor runs from 85* up to about 125* water temp. It is transferring heat through 3/4 solid wood flooring so I keep actual floor temps at 85* or less.
I wired up the Windhager for constant demand and simply tee'd it into the supply between the boiler and the heating pumps and ran the return back from the main system return piping. Took about 20 minutes to make the connection in the basement. (extra tee's with valves under them make for easy experimenting)
I played with the setpoint on the Windhager and wound up leaving it at 158* after beginning at max temp. When it was mild outside, (temp above 25-30*) I would routinely see the water temp creep up past the setpoint even with the boiler modulated down to 30%. The highest temp I observed was about 175* for any duration of time. If it stayed there for more than 15 minutes or so the burner would simply cycle off while the pump on it continued to bring heat into the system.
Installed on a system like this I can see absolutely no need for a buffer tank or storage of any kind.
As noted above though, the devil himself lives in the details and if you connected the same boiler to an on/off load, you would probably enjoy some benefit by the use of a buffer tank. Something as small at 40-50 gallons would be sufficient because of the modulation and the fairly quick shut down start up time frame.
A better way in my book, assuming a solid heat load calc has been done, woud be to size the boiler appropriately and then use a panel or cast rion rad with a TRV on it in a basement. Any heat from a rad in a space like that is not "lost" and it would serve the same purpose of limiting the number of on/off cycles.
If a person were running a 21KW (71,000btu) BioWIN for example, a "load" of only 10,000 would prevent the boiler from cycling off in all but the most mild temps. If I have the boiler yet next winter I will be putting a CI rad in the basement to see how that affects operation.
In the case of a heating design that is truly all or nothing such as a forced air (yuck) system a person has two options. Install a small buffer tank or adding a small "constant" load. In most cases I would recommend adding a small rad to dissipate a little heat while not actually wasting it. This would be far simpler and less expensive in terms of up front cost and it would also reduce the on/off cycles. IIRC, in talking with the Windhager people they recommended shooting for a cycle/hours run ratio of no less than 2:1 so a person would have to consider a number of variables when designing the system layout.
Hope that clarifies things a bit.
