Water from our boiler is less and less often reaching the last four loops on the 17-circuit manifold, and I was wondering whether the problem is the pump or the water supply or something else (for example, two of the rads in the house are leaking near the bleeder valve, and tightening things accomplishes nothing; replacement of parts has had to wait until the new year).

Background: We have just moved into a ~2200 sq-ft, 75-year-old house with a very recently updated hydronic system: 2-year-old boiler (RBI Spectrum), pex tubing, supply and return manifolds, each with 17 inlets/outlets: the inlet and outlet nearest the boiler are capped; there are 14 rads (2 years old), and I presume each has its own loop; there is a loop for a subsystem of three radiant-heated floors (4 sub-loops) with its own 2-speed pump; I'm darned if I know what the unaccounted-for loop does. The previous owner did not, however, replace the auto-feed, the circulator pump, the pressure relief valve or expansion tank when he replaced the boiler. We replaced the pressure relief valve and expansion tank before we moved in because they were not working.

The radiant-floor loop is meant to be fed from both a solar thermal system in the garage and the boiler, but the primary solar loop is not working properly, so we've closed the secondary solar valves on both the return and supply lines for the radiant floor subsystem, and now that subsystem is being supplied by, and returning water to, the boiler alone. The inlet & outlet for the radiant-floor subsystem are third from the end on the supply and return manifolds.

When we first started the rad/boiler system up for testing and maintenance, there were three rads that did not get warm: 1 of the 2 in the basement, the wild rad on the main floor, and one small rad on the 2nd floor. Eventually, as we let the boiler run longer, all but the small rad on the second floor got hot (That one heated up _once_ recently. After we'd had some trouble with the system and called in a repair person on Dec. 20, the rad got hot when the system was turned back on, but the little rad stayed hot only a few hours).

Two days after that, Dec. 22, the pilot light went out and could not be relit, and we had a wait of several hours before someone else from the same HVAC company could come to replace the thermocouple. In the meantime, water began to drip out of the pressure relief valve, and from around the valve at the top of the expansion tank, so we turned off the water supply to the boiler. When the tech came, he said that the leaks happened because the auto-feed mechanism is old and unreliable and needs changing very soon. I could not tell you now whether/how much the tech turned the water supply to the boiler back on; I know it is not fully on, because there is room to move the valve to the left. Because of his worries about the unreliability of the autofeed, I'm chary of fully opening up the valve lest there be a catastrophic failure.

In any case, now, both of the basement rads (the originally 'cold' one _and_ the leaking one) as well as the wild rad frequently go/stay cold when the boiler comes on and the other rads heat up. In fact, fairly frequently -- I'd say 'more often than not' lately -- the last four inlets/outlets on the manifolds do not get warm, so the radiant-floor loop is gettin no hot water either. Aside from the perplexing question of how four loops at the end of the manifold correlate with four rads _and_ the radiant-floor loop, I would like to know why hot water does not regularly reach the four circuits at the end of the manifold. If I open the bleeder valve on all the rads, water comes out immediately.

The repair person we had in on Dec. 20 to service the rads said that the little rad on the 2nd floor doesn't get hot because the pump is not big enough for the job (it's a Taco 007). He also said we ought to get a 24-hour circulating pump so that the basement would not get so cold when the boiler was not running. The Husband (a pretty savvy DIYer who, nonetheless, admits he doesn't know anything about boiler systems) fears the HVAC guy is trying to sell us something we don't need (The Husband is also something of a penny-pincher).

So, why is the little rad on the 2nd floor consistently cold, _and_ why are the last four circuits on the manifold increasingly frequently cold? A weakling of a pump? Not enough water? Leaking rads? A combination?

Also, if I open up the water valve to the boiler very slowly, will I avoid hastening the auto-feed's failure?

Thank you for your patience in reading this.