We are renovating our basement and need to make a decision on heating systems. We live outside of Boston and the space is about 650 square feet. The existing concrete floor needs to be jackhammered and replaced because it so uneven (story for another day), so installation of radiant cost are minimized when pouring new floors. We already have radiant heat in the house, so adaption from the natural gas hot water heater should not be difficult. However, installing hot water baseboard should not be difficult either. Knowing that radiant is less expensive to operate but more expensive to install, what factors does one use to decide on radiant head or baseboard heat in the basement?

Radiant heat in a basement slab is easy especially if you are removing the old slab and have a boiler delivering low temp h2o now.
When you remove the slab and prep for the new you will want to provide 2 inch for rigid insulation under the new slab.
Pipe can then be tied to re-bar or wire mat or stapled directly to the 2 inch foam.
If you have in floor heat now check the delivered water temp, odds are you can hit the slab with the same temp water if you are lucky.
pipe on a 9 inch pattern will be more than adequate for most basements, some will argue for less. pipe is cheep. a 900' role will cover the whole area.
Dan

Two inches is on the way-skimpy side for basement-slab radiant in MA. You only get one bite at the apple so make it a big one. XPS will degrate to the same R value as EPS well within the lifespan of the slab and does more than 100x the environmental damage (mostly from the high global warming potential blowing agents used). To be nice to the planet, use EPS. At 2" you're looking at a lifecycle-averaged performance of about R8 for either, which has a reasonable financial rationale for a slab that ISN'T being used for radiation. See Table 2, p.10 of this document- the row for US Climate zone 5 (where Boston is located):

http://www.buildingscience.com/documents/reports/rr-1005-building-america-high-r-value-high-performance-residential-buildings-all-climate-zones

For a radiant slab in Boston, consider 3" (R12) the starting point, and if it's a lossier than average basement (requiring a higher average slab temp) 4" isn't necessarily going to be a waste.

Type-II EPS (1.5lbs nominal density) has more than adequate compressive strength for handling residential loads under a 3-4" slab, but if the inspector insists on a higher psi rating, Type-IX (2lb density) has the same ratings as XPS.

Under no circumstances should you use polyiso (any density) under slabs, since it is somewhat hygroscopic and can become waterlogged over time.

Clearly baseboard and only 2" of sub-slab foam will be cheaper to install, but it has nowhere near the comfort factor of a radiant slab.

You will first need to know the total basement heat loss and the exposed floor heat loss (with your chosen floor insulation R-value as Dana suggested) to determine the design parameters for a hydronic heated floor.

Assuming a conservative 10' leader length from each circuit to reach the manifold station, you will need four 237' circuits (947' total) for 650 SF at a 9" spacing. Assuming a non-conservative 0' leader length from each circuit to reach the manifold station, you will need three 289' circuits (867' total) for 650 SF at a 9" spacing. In case this isn't obvious, you should not use one 867' circuit because of pipe friction and heat transfer limitations. Please feel free to run heat loss analysis and hydronic floor design numbers yourself using the free software on our website to get a better sense of what will be required.

Whether you go with floor or baseboard heating could only be determined with a detailed acquisition cost and operational cost analysis, but I would also predict baseboard to be cheaper unless perhaps you are doing this project yourself. I vote for comfort too and if you have to pour a new slab anyhow, purchasing and placing 0.5" PEX is easy and cheap like Dan indicated.

We make a living tearing out concrete, insulating and radiating old basement floors. We do not install fin-tube since it is the lowest of the low in hydronic heating, comfort and efficiency. Most basements here in Minneapolis can be heated with a third the tubing suggested. One of the most important considerations after insulating below the slab and foaming the walls is that the floors will rarely "feel" warm. But there is no question as to the efficiency immediate and sustained.

Always tube a slab.

Always tube a slab.

It would be interesting to compare the additional cost of aluminum plates to the operational cost savings of being able to do thermostat setback. Ie, a high mass radiator means overshoots, undershoots and generally prevents the use of thermostat setback. So I'd pay something more to not have it.

high mass does not mean over/undershoots. that's a water temperature control problem, not an inherent problem of the mass, barring extreme situations with extremely thick slabs. it's marginally worse in solar gain situations.

quite certainly the cost differential you are looking for would, in any well insulated home, not be nearly enough to drive a decision to install aluminum over the slab. exceptions might be made only in cases where the system oscillates between heat/no heat modes regularly (very mild climates, partially heated spaces like part time workshops, etc) or in cases where the mass is extreme (very thick slabs).

for any 4-6" slab in a regular heating climate in a regular heated space, pipe the slab!

thermostat setbacks are a minor advantage in well insulated conditions. they are primarily of advantage in leaky homes with oversized heating systems. That happens to describe an awful lot of existing buildings in the US but it shouldn't describe any building getting radiant because before you did the radiant you fixed your envelope, right?

that's a water temperature control problem

It's a change output rapidly problem. Slabs don't do that.

Like most things, setback savings (in btu) go down as the the house gets better insulated . The savings also go up or down depending on schedule, outdoor temperature and what other mass you have. A low interior mass office building is a good example. Reference Siegenthaler, who discusses setback being ineffective only in "very well insulated buildings with high thermal mass".

well insulated buildings don't have to change their output rapidly. Even poorly insulated buildings rarely have to change output rapidly. output requirements don't typically fluctuate that rapidly other than in passive solar situations.

Setback in particular is a savings that only occurs if you can actually drop your inside temperature appreciably over a night. well insulated homes don't do this. thus there isn't much savings available, and that is why us radiant guys typically pursue lower water temperatures instead of setback-ability. Plus, setback is a comfort crippler when you come out... you then have to reheat all the mass in the home before you hit an equivalent comfort level for a given thermostat setting. Radiant is better at that than forced air, but still, it takes time. I have recently estimated savings per degree of continuous thermostat setting as a 5% reduction in yearly loadhere in maine for any home. doing it overnight for eight hours would be, at best, one third of that then, and that would be in a zero mass/instant loss-recovery setback. In practice, with the time to lose the heat and regain the heat, you are unlikely to crack 1% savings in a modern home for an eight hour overnight setback. and you'll be less comfortable as a result in most cases.

there are other strategies that can be deployed like spot heat to raise local MRTs to comfortable levels even while the space is cool or what have you. but they have pretty limited application IMHO.

We also always recommending putting tube in a new slab to provision for any future desired hydronic heating option. Tube and placing it then is relatively easy/cheap compared to having to add a tube/plate above-floor assembly later.

Yes, a non-conservative 0’ leader length from each circuit to the manifold station and two 163’ circuits (325’ total) for 650 SF at a 24” spacing will get the job done if your bare feet don’t mind the larger floor temp gradients and if you don’t mind having a supply temp that will be about 10 degrees F higher than using the 9” spacing. Of course, you really can’t spec the spacing until you know the BTU requirements and how the room will actually be used.

There is certainly an operational savings that may be gained using temp setback, but you also have the added acquisition expense of needing a larger heat source to allow subsequent raising of the building temp in a reasonable amount of time (and this can vary significantly depending on the heat capacity of the furnishings in the building) and there may be an operational efficiency loss associated with having to use this larger heat source. High mass radiator overshooting and undershooting can be easily mitigated with a good control design. I concur with Rob.

output requirements don't typically fluctuate that rapidly other than in passive solar situations.

As in any closed door room with a east, south or west window? I have standard rooms, not particularly designed for solar gain, that go from substantial heat needed to none needed in minutes. Once the sun starts coming it, an 80F slab is only making things worse (ie, overshoot). I'd like to hear more about the control design that prevents this.

I have recently estimated savings per degree of continuous thermostat setting as a 5% reduction in yearly loadhere in maine for any home. doing it overnight for eight hours would be, at best, one third of that then, and that would be in a zero mass/instant loss-recovery setback. In practice, with the time to lose the heat and regain the heat, you are unlikely to crack 1% savings in a modern home for an eight hour overnight setback

The studies (admittedly somewhat old), usually find around 12%. Your figures pretty much support that - 5% per degree of setback times 1/3 of the day, times 10 degrees of setback - that's 17%, then adjust for slow response time (which depends on the house and the outside temperature). But somehow you knocked 17% down to 1% (basically saying that the house doesn't change temperature). Low mass houses do cool down/heat up when you turn the hvac off - and it doesn't take all night.

the sun doesn't, for most people and the vast majority of homes, go from zero to full strength immediately. a morning passive solar overshoot is about the only thing hard to accommodate because it's the fastest shift there is, so if you have a lot of east glass in a high mass room that is definitely an issue if overshoot in that case is a problem. Of course, you'll overshoot in that case in a low mass scenario as well, but it won't be as severe. In that case you want high mass that perhaps is not already hot. Add panel radiators, not low mass radiant subfloor products.

south and west solar events don't sneak up on you, typically. You've had warming occur for hours beforehand and your heat demand probably ended a long time ago as well. and you NEVER go from "substantial" heat needed to "none" in minutes in any other event I can think of short of an intermittent internal gain like a fast heating woodstove though again USUALLY you are starting a fire in a chilly room not one that is already at full room temp. You go from "none" to "some" when the sun goes away, maybe, and that can cause an undershoot eventually if you are not well controlled... and that's again a control issue, not an insurmountable mass problem, a floor sensor can solve it.

but you have heat in the room, floor, walls, ceiling, furniture, etc that has to dissipate before you actually feel cold even when the sun goes down. the presence of a window isn't enough to make room temps fall automagically. it has to be a lot of window. thus my passive solar comment.

You find me a decent house that drops ten degrees for eight hours a night. I do overnight setbacks in my shop and I can, if I try, get a 5 degree drop in my slab level in about 12 hours of setback which is a 3 side exposed floor with 3 doors and 2 garage doors and two windows.... hardly a paragon of perfection for low loads... in my upper low mass level I don't get near ten degrees a night either. You are just throwing out numbers that have no basis in reality. a 5 degree setback is deep for any decent construction these days except in that cathedral great room with the window wall...

and you (and most setback proponents) are just ignoring the fact that you're still colder when you come out of recovery at an equivalent thermostat setting because of the now cold surface temps of everything in the home. The colder you let it go at night, the chillier it is in the morning when you come back, because that mass takes time to reheat. If you're ok being chilly, then just turn the thermostat down a few degrees in the first place and KEEP IT THERE. You'll save more. I have repeatedly diagnosed severe comfort issues in air heated homes with internal mass (slab on grade) as an overuse of setback... they stop setback and suddenly (well, a couple of days later, usually) the home is actually comfortable with no other changes. It's amazing.

but you are correct in that my "you are unlikely to crack 1%" comment was in error. You are unlikely to crack 1% PER DEGREE of thermostat setback, is what I should have said, I apologize. 1.5% (roundish number, we're estimating) per degree of actual AVERAGE temperature reduction over the 8 hours, or typically about half that for a given setback setting if you assume even basically equal heat/cool times... maybe rounding to a nice 1% if you presume the system can provide heat faster than loss, but if you are doing that, then your heat source or your water temp is over-designed and you're sacrificing efficiency elsewhere... in greater amounts.

so substantively I still stand by my statement. in modern construction, absent a window wall or passive solar design situation (much more severe than "a" window), mass radiant is a control proposition that is easily solved with water temperature and/or slab mass temperature control. that is something low mass systems may not need, and it doesn't require someone to know what they are doing to get better results to go low mass there, but it's not a problem inherent with mass. Mass can be controlled. Temp swings in modern construction should not be extreme or rapid in the vast majority of situations. I have many hundreds of radiant systems out there that can attest to this basic fact...

Rob is right.

Like "raising the tube". It sound like a good idea until you do the math. Basements have always been a relatively stable thermal environments. Dramatic ambient temperature swings do not happen, hence no need for rapid recovery nor sloughing off unwanted heat. As Rob asserts, it is very easy to control any thermal mass with reset water temperature. Much like a 48,000 lbs semi-trailer is controlled with cruise control and common sense.

Rob is indeed right, but don't forget to also include the efficiency hit for having to operate a heat source that may have to be sized significantly larger than required for normal operation just to deal with this large temp setback recovery. Folks often forget that "minor detail" when advocating large temp setback.

Jonr, relative to wanting to hear more about control designs... A "model-follower" control system design approach will mitigate high mass radiator overshooting and undershooting for a building with an integrated passive solar and hydronic floor heating system. The "model" in this case is a math model of the predicted zone heat loss that will occur given the actual/forecast outdoor temp and the desired zone indoor comfort level schedule (which you can think loosely as the desired zone indoor temp even though actual indoor temp is not a primary control feedback parameter for this control system design approach) and adjusting for the predicted heat gain that will occur given actual/forecast solar irradiance and any scheduled supplemental heating.

Given the predicted adjusted heat loss for this zone, the model predicts the required upward heat gain that must be provided by the hydronic floor heating system and the required slab temp that will provide this upward heat gain. Having this predictive model allows the control system to slowly and successfully "follow" this model, minimizes encountering large and unknown variances that could exceed the control system authority and reaction time, and keeps the slab temp precisely where it needs to be in order to always maintain the desired indoor comfort level schedule for the zone. While you still can't perform a large temp setback like you can with a low mass radiator, you can successfully perform enough of a setback to see reduced heating cost.

This is of course a very simple description of a model-follower control system. Advanced math and programming skills are required to successfully design and calibrate the control system. However, this is the best approach we have found for both maintaining desired indoor comfort level and minimizing heating system cost for a building with an integrated passive solar and hydronic floor heating system.

"Rob is indeed right, but don't forget to also include the efficiency hit for having to operate a heat source that may have to be sized significantly larger than required for normal operation just to deal with this large temp setback recovery. Folks often forget that "minor detail" when advocating large temp setback."

Like the other 98% of the country that installs and swears at/by scorched-air.

We are better than that.

Set-back should be left to those who don't really know or care about real comfort, e.g. the power companies and their tin-headed cohorts.

Oh...was that brash?...hehhehehee

the sun doesn't, for most people and the vast majority of homes, go from zero to full strength immediately

It pretty much does, frequently. Cloudy, then it suddenly clears up. Usually combined with increasing outdoor temps at the same time.

You go from "none" to "some" when the sun goes away, maybe, and that can cause an undershoot eventually if you are not well controlled... and that's again a control issue, not an insurmountable mass problem, a floor sensor can solve it

No floor sensor or control will make the slab change temperature quickly. In many such cases, you will have undershoot.

The "model" in this case is a math model of the predicted zone heat loss that will occur given the actual/forecast outdoor temp

Weather forecasting, particularly partial cloud cover is horribly inaccurate. So mitigate yes, eliminate no. Even if there was an accurate prediction that heat load would change from requiring an 80F slab at 9am to a 70F slab at 9:30am, how would a control system make this happen?

A non exposed basement heat load not changing quickly and recovery from setback needing extra capacity - I agree with these points. Luckily, a control can easily adjust setback to account for extra capacity - which all systems have most of the time (all conditions less than design temps). Use of this costs nothing in efficiency (multi-stage heat pumps being the exception). Cooler while sleeping is more comfortable. Brash people often miss these points.

I don't like setback in bathrooms. But they are small and so inexpensive to keep fully heated all the time.

"Weather forecasting, particularly cloud cover is horribly inaccurate. So mitigate yes, eliminate no. Even if there was an accurate prediction that heat load would change from requiring an 80F slab at 9am to a 70F slab at 9:30am, how would a control system make this happen?"

Actually, determining what the outdoor temp and solar irradiance will be 8 hours in advance is easily accomplished, is extremely accurate, and is all that is required to provide sufficient control system lead time. There would never be a situation in a properly designed and constructed building using a model-follower control approach where you would ever need to change the slab temp by anything like 10F in 30 minutes...or even 1F in 60 minutes. The design objective of an integrated passive solar and hydronic floor heating system is to maintain a desired comfort level at all times while minimizing heating system operational cost...and NOT change the slab temp any more than is absolutely necessary to accomplish this.

Jonr, you are making very uneducated assertions. If I have a floor sensor in your slab and I don't let it go below 70 degrees, you will not have an undershoot in a slab zone from a dissapation of solar gain. Ever. When a sun goes behind a cloud, or sets, your room temperature does not change immediately. the LOAD changes, sure, but not the TEMPERATURE... not immediately, and in modern construction, not that quickly in most cases. because you have embodied energy in ALL homes that mitigates internal temperature swings. You have some massless conception of indoor comfort that quite simply is not real or accurate. If your room temp DOES change that fast when cooling down, then you do not have good modern construction or you're in the window wall great room I keep mentioning. it if heats up that fast, then you're in an exceptional solar gain situation.

the ONLY conditions we cannot accommodate with an off the shelf control system in a mass home are:

-extreme morning solar gain
-extreme solar gain that overwhelms ANY system
-fast heating heat sources like woodstoves that are triggerred when the room is already warm. NOT FAST COOLING SOURCES because the ROOM STORES ENERGY TOO. only heat up is a problem we can't fix quite so definitively.
-VERY massive slabs (more than 6" thick is my informal definition)

Anything else I can control within +/- 2 degrees with indoor temperature feedback controls and floor sensors which is a range almost everyone considers quite adequate. Outdoor reset is less responsive when heating up than indoor feedback but still a good PID or PWM thermostat should keep it locked in. I know because I do it, all the time. No weather prediction needed. If you want to get into weather prediction, that might be able to address the solar gain preparation for the few cases we can't handle otherwise. But residentially it would very rarely be affordable or worthwhile to attempt that and it is only a consideration in very rare circumstances.

Yabbut, jonr lives in a tent, the only substantial thermal mass being jon and his dog! :-)

But seriously, all houses with insulation & windows with intact glass have substantial thermal mass and even with code min R/U, a significant time constant to smooth out the short term changes in load. A 4" slab doesn't have such a huge amount of stored heat that it can't be managed for the intermediate term swings in load, even with a comparatively low mass house around it. (And in an insulated basement the loads are so low and slow changing that it hardly matters.) The worst-case scenarios are solariums of over-glazed passive solar homes, where the U-factors are high, and so is the solar gain, and using the thermal mass storage of the floor as radiation can indeed result in pretty big temperature swings in that room, but even that can be tamed to some extent. (Unlike clouds, the timing of dawn is fairly predictable, even if the exact solar gain at that time isn't.)

If I have a floor sensor in your slab and I don't let it go below 70 degrees, you will not have an undershoot in a slab zone from a dissapation of solar gain. Ever.

Evidently you claim that a 70F slab (and other 70F mass) will keep a room with heat loss at 70F without undershoot. Study some basic physics (or do measurements) and you will see that this is always WRONG (over any time frame). All you can rationally argue is how much undershoot there will be before the slab can rise to the temperature it needs to be (many hours). -2F can be correct, as can -5F; as you seem to agree, it depends on the amount of room mass, slab mass, the heat loss rate, floor coverings, how oversized the heat source is, how outdoor reset is set, etc.

... not that quickly in most cases

Now you are getting more reasonable. As I said, high mass radiators cause overshoots and/or undershoots. When exactly are they big enough to be objectionable (yes, this occurs) - I haven't commented on that.