The following is a re-posting of a Google Docs document that will be subject to constant correction and update from reader comments.
https://docs.google.com/document/d/1o3j4ayBkhDhq8U9HVDwYujwKQCOQLpANUYvNFKFWfco/edit?usp=sharing
Better HVAC Air Ducts In Our Homes
The following message is sent to the Duct Design leadership at ACCA and to ASHRAE Ducts Design Committee colleagues, TC 5.2.
From my PhD-level work forty to fifty years ago, with hydraulics of pressurized water nuclear power generators, I have an understanding of HVAC duct design with direct application of Bernoulli Principle. This understanding contradicts ACCA Manual D design of residential HVAC ducts, where there is no usage even of the word Bernoulli. By ACCA methods, nothing is seen as wrong in silly long too-small ducts that sprout Medusa-like from beloved D-boxes and assorted Supply Trunks. Absent Bernoulli math, air is somehow transported as if it were no-energy toothpaste. Static pressures change and diminish for reasons little-known, not controllable. Manual D pros say: Just squeeze enough at the blower,to push the needed flow. Flow violence and attendant inefficiencies are only a noise problem. Ignorance is revealed in non-scientific phrases total static pressure and total external static pressure.
With Bernoulli energy balances, we have spectral local values of static pressure and kinetic head, additive to total pressure..Energy losses attendant to poor design are more-usefully computed, than measured. Better than computing penalties, use common sense to avoid them. Let the controlling flow resistance in any path be at the register discharge to ambient. Find folly always in trying to fix a mess with use of dampers. Resize any register and its approaching duct, to deal with complaints.
Please see a best demonstration of my duct construction and analysis methods in one home:
Duct Hydraulic Analysis, Koempel
Koempel Job Photo Album
Blog Post An Attic Ladder Installed Diagonal to Attic Floor Framing
Blog Post Flawed Measurement Of R-Value With A Certainteed InsulSafe4 Gage
I have several similar achievements to share:
Chamberlain-Mann Crawl Space Ducts Hydraulic Analysis
Chamberlain-Mann Crawl Space Job Photos
Waters HVAC Ducts Plan
Waters Attics All Photos
Duct Hydraulic Analysis, Leet Rental
Blog Post, Leet Rental Crawl Space
Leet Rental CS Overhaul
Leet Home Better Furnace Ducts
HVAC Ducts Hydraulic Analysis_Meyer
Job Album, Meyer Attic
Find more duct redesign from found cheap industry expectations. Employ shop-built fittings, often lined with insulation. No leaky, inefficient on-the-job hacks to just make connections. Ducts are buried as much as possible, never placed carelessly, obstructing access.
Blog Post: Best Placement of an In-Attic Gas Furnace, Among Roof Trusses
Blog Post: More Furnace Plena and Flex Duct Quality Work
Blog Post: Following ACEEE Blog, Furnace Fans As Energy Hog
Blog Post: Better In-Attic Furnace Ducting
Blog Post: Steel Ducts Heat Capacity, Forced Hot Air Heat
My insight in all of this is with confidence from my work in the nuclear power industry where a key achievement was in the reversal of a management decision that the cost of reactor pressure vessels should be reduced by making the inlet and outlet nozzle forgings abrupt, sharp-edged, no chamfering. A million dollar scale-model flow test under my direction proved that flow is then with damaging violence. Such silliness.
Please see that these contributions are extraordinary. All measures taken for energy efficiency are affordable. The wonderful sheet metal shop that enables my work shall wish for more customers like me. At present the cleverness I offer to customers has insufficient demand.
ACCA Manual D does not offer scientific best methods. Together ASHRAE and ACCA must offer better guidance. I want to help.
Phillip Norman <pjnorman@gmail.com>
1764 Bonniebrae Drive
Lake Oswego, Oregon 97034
503-255-4350
My Customer Meyer:
Please visualize the defect of D-box work imagined as no-cost squeezing of toothpaste. The discharge of a blower over a plenum area is quite directional. Noisy turbulent eddies form if flow does not remain guided in consistent direction. In a D-box, the orderly kinetic energy fan-generated, is not converted in efficient buildup of static pressure to drive lateral flows through the resistive and leaky sharp-edged hacks. Chaos has a cost. energy-wasting large eddies uselessly elevate the blower head.
Instead, use common sense to create great value in a very tall attic space.
(11/24/2017)
Let the furnace and ducting become non-obstructive. With very little energy loss, silently turn the developed air flow 180°. Bury all ducts under a maximum of batt insulation covered and protected air-tight. Retain one awful feature of the queer found installation: There is no return air filter!
When, later, an air conditioning heat exchanger was added, there still was no way to add a filter box.
Here is a composite photo of the dark and daunting found-conditions.
All attic access is obstructed. Even the small OSB furnace floor is blocked. One must dangerously duck under 2x4 truss bracing. All of this is of insult to USA attitudes toward attics. Don't go there. Let them be booby-trapped. Let all attic work be delegated to newby unfortunates, without inspection and thoughtful supervision, never to be improved.
Now talk about the new possibilities.
Look down upon efficiently-dividing paths below the 180° turning plenum. See guided concentrating flow continue at steady velocity, quietly and efficiently. Flow is quite turbulent, but with very small energy-blasting eddies. Regardless of frenzied froth, flow wants to be guided, and absent guidance behaves wildly and noisily. All needed turns must be efficient, at lowest-economical velocity, all the way to a controlling flow resistance at discharge in registers.
Where flow may divide in wyes, each path to register discharge may be individually summed in a Bernoulli balancing analysis. Where the dominant resistance is at each register, register velocities will be equal as obvious good design. The ONLY place to throttle flow is in a register. Control the service to a room in the sizing of its register and approaching duct. Keep duct sizes as large as possible except in a consistent final length, say of six feet, at each register.
Here see effort to reduce steel-duct swung thermal mass by internal insulative ceramic coating, that does matter.
Let the planned flows be charted for analysis, in diagrams such as this for the plenum above:
Here is the completed ductwork at the furnace:
None of this simple common sense is the Medusa-teaching of ACCA Manual D.
The convenient found storage space in this beautiful attic now super-insulated, is quite valuable.
At November 2024 I share news of yet-another project eliminating a leaky, thoughtless hack-in D-box.
One Before photo:
See common shabby dirtiness in a Medusa of pipes register-sized full-length, blown air tripping over and leaking past barely-capturing tabs. Nothing stably guides the flow into a pipe.
Yet, reduce the flow resistance to a simple
Loss coefficients are dimensionless multipliers upon velocity head (air pressure of motion), at bulk velocity in the pipe approaching the register. A bulk velocity is the ratio of a measurable flow rate, and pipe area.
Here is the shop-built replacement supply ducts top-box of this furnace:
Let developed flow remain guided and silent. Here nine home registers are geographically collected to four equal outlet paths. Nothing in ACCA Manual D recommends this. This 21" square 12" tall and with 9" outlets might become a common standard supply box. Flow progresses through sequences of wyes, always maximizing pipe size such that the discharge register is the main flow resistance. Want the total loss coefficient, K in each path adjusted to final duct velocity, to be little more than 2. It is easy to do. From my cited hydraulic analysis examples, see that path K with careless design is commonly more than double this. Undersized portions of any flow path add to troubles.
Well-guided flows are equally divided in the four 9" diameter supply outlets, branching to individual registers through wyes and straight pipes of maximum size. The goal always is that the dominant flow resistance in each register path is at that register and a minimum of approach pipe length at register size. Where most pipe length is upsized by at least one inch of diameter, any register might be upsized as-needed, with a minimum of expense.
Sensible ducts are as large as economically possible, reducing progressively beyond each wye branch as in any living circulatory system, down to a controlling squeeze through a final capillary at each register.
But, is the register always that squeeze? Here and often, maximun air velocities and pressure losses are atop the furnace in the supply starter ducts. Space upon a supply distribution plenum atop the furnace is limited. Even with Medusa intent, a supply-duct starter often feeds several registers.
Anytime a wye is employed, expect that the smaller branches will be relatively small, with space a linear limit, and area larger by square of diameter. Pressure losses in the starter duct are higher by the square of duct area ratio:
Say each starter duct has a loss coefficient at local duct bulk velocity head, of 0.5. Tabulate that loss adjusted to register duct velocity head, i.e. multiplied by the square of the area ratio:
Where the local loss coefficient is much smaller than in a hacked-in D-Box, the k-value at register velocity head remains small, and the 9" starters are adequate.
With 9" or 10" starters, local velocity head is greater than that approaching registers. Air motion through the furnace and heat exchangers is energetic. The motion must be guided and efficient.
There is real, substantial cost of all inefficiency. We can afford engineered and shop-built best components. Reject determination by fabrication cost and pennies pocketed by a builder.. Lifetime operating cost is the concern.
How now, do I inspire others to follow my example of thoughtful HVAC duct design? It seems that ASHRAE and ACCA will not help. Look for an evident likely ally, a respected believer, to weigh in. Conduct this Google search: blower energy cost savings if I reduce flow resistances by half . Find this excellent simple and scientific post by Donaldson Filtration Solutions, Donaldson Company, a global manufacturer based in Minneapolis:
See that piping flow resistance reductions are simply additive to that of a new money-saving and heat exchangers-preserving much bigger air filter. Furnace filters and heat exchangers are all additive resistances. There is a penalty in having two in-series heat exchangers in a forced-air gas furnace with air conditioning. Avoid that penalty in conversion to an electric heat pump for this home someday.
.
