HVAC design

hello there. currently, i'm working on electrical, plumbing, and hvac systems. i'm going to focus on the hvac system in this post.

the system

i selected & purchased a system - it is a multizone split heatpump system made by fujitsu. it comes with one outdoor unit, model aou24rml1, and two indoor units, model aru12rml. this is considered a two-ton system. the indoor units are both blowers, and each comes with a wall-mounted remote.

the system can be installed as two separately controlled zones or two blowers in a single zone, controlled at one or two points. there are also cool accessories like wireless remotes that could be purchased, but i'm sticking with the basics ;)

the outdoor unit is smaller than a typical outdoor heatpump, and it employs a variable-speed compressor - 'inverter' type. this allows the system to operate at very high efficiencies during part load conditions, which makes it very efficient! good stuff.

the only downside is that this type of system is uncommon here, and if there are ever problems with the system it will be more problematic to fix, as the owner will likely need to find a service company that deals with fujitsu systems. my thinking is, hvac equipment efficiency is constantly going up, and by the time there is a problem with the system it will probably be obsolete and the service company will probably recommend replacement.

following this train of thought, i would like to make the existing system less problematic to upgrade/replace in the future. specifically, i'm focusing on designing the ductwork so it can be adapted to a more traditional single blower system. this should save the cost of conversion in the future.

however, i don't want to make any significant sacrifices in the efficiency of this system to meet that end. so, it requires some careful consideration in ductwork planning.

ductwork design

there are various methods of designing ductwork, from hand methods to computerized solutions. i've looked over these and considered the physics from the blower down through the registers, from efficiency & comfort considerations. i haven't found a clear method spelled out that satisfied me, but i have arrived at a process that i believe is optimal to follow - as it seems to optimize all considerations.

i did alot of reading and thinking to come up with this, and i'm providing it as a useful resource for anyone who might come across this post on the web. take it for what it's worth. it is summarized thus:

velocity = manuf(CFM, throw)
boot_size = area_to_diameter(CFM/velocity)

CFMs at each outlet are determined by heating/cooling needs (typ. a manual j calc) and the CFMs available by the blower. the number of outlets in each room are adjustable to arrive at a typical-range CFM at the register to give good air distribution.

throw at each outlet is dependent on the room dimensions and the register location (read up on ADPI for more details on this.) for typical baseboard registers with spread vanes, throw should be 0.7 x distance to far wall (gives ft/s.)

velocity at each outlet is found by referencing manufacturer data sheets of the selected register/grille. it will depend on a combination of CFM and desired throw.

the outlet/boot beneath the floor register/grille is sized by dividing CFM by the velocity. this gives the cross-sectional area of your outlet, and using a = pi x r^2, we find diameter = 2 x square_root(area/pi).

---

all supply and return branches run from the blower to conditioned space. the pressure at the blower is the same for all branches because it's in the same place (obviously) and the pressure in conditioned space should be the same (so use passive ducts between rooms if necessary.) thus, we can see why the total pressure drop of each branch will be equal.

the total pressure drop of a branch is the total equivalent length (TEL) times the friction rate (FR). TEL x FR of each branch should be equal. if they aren't, the delivered CFMs will deviate from that expected.

TEL of each branch is a sum of duct lengths and the equivalent length of all fittings. equivalent lengths can be found in data tables. FR is dependent on CFMs, duct diameter, and duct smoothness.

a 'ductulator' program is used to select an appropriate diameter for given CFMs and desired FR. if some section of ductwork is a size not consistent with your desired FR, you can calculate the pressure drop of that section only with it's TEL x actual FR, and add that to the TEL x FR of the rest of the ductwork.

for even more accuracy, instead of using equivalent lengths of fittings, the friction loss at a fitting can be calculated using a coefficient in conjunction with the velocity pressure (which is dependent on velocity at the fitting) - and these coefficients are found in data tables for the fittings. such a method would add the fitting pressure loss to ductwork length x FR

And so you can now size all of the ductwork.