DEV Community: Evgenii Konkin

Enthalpy vs Temperature: Why Sensible Load Alone Undersizes Cooling Coils

Evgenii Konkin — Fri, 25 Sep 2026 02:59:52 +0000

Temperature tells you half the cooling story. Enthalpy tells you the whole thing.

What enthalpy captures that temperature misses

Sensible heat = heat from air temperature (what a thermometer reads)
Latent heat   = heat from moisture content (invisible, but real energy)
Enthalpy      = sensible + latent combined (BTU/lb of dry air)

The formula:

h = 0.24 × T + W × (1061 + 0.444 × T)

Where:

h = enthalpy, BTU/lb
T = dry-bulb temperature, °F
W = humidity ratio, lb moisture / lb dry air

Same temperature, different enthalpy

Two air conditions at 80°F:

Condition A: 80°F, 40% RH
  Humidity ratio: 62 gr/lb
  Enthalpy: 28.5 BTU/lb

Condition B: 80°F, 70% RH
  Humidity ratio: 110 gr/lb
  Enthalpy: 37.8 BTU/lb

Same temperature. 33% more total heat in Condition B. A coil sized for A cannot handle B.

The coil load comparison

5,000 CFM outdoor air cooled from 95°F/75°F-wb to 55°F/54°F-wb:

Sensible-only calculation:

Q_sensible = 1.08 × CFM × ΔT
Q_sensible = 1.08 × 5,000 × (95 − 55)
Q_sensible = 216,000 BTU/hr = 18.0 tons

Enthalpy-based calculation (total load):

Q_total = 4.5 × CFM × (h_in − h_out)
Q_total = 4.5 × 5,000 × (38.5 − 22.7)
Q_total = 355,500 BTU/hr = 29.6 tons

The gap:

Sensible:  18.0 tons (61%)
Latent:    11.6 tons (39%)
Total:     29.6 tons (100%)

Sizing by temperature alone: 18 tons. Actual need: 29.6 tons. The coil is 39% undersized.

When latent load dominates

Climate              Latent %    Enthalpy matters?
──────────────────────────────────────────────────
Phoenix (hot-dry)    5–15%       minimal
Denver (moderate)    10–20%      moderate
Atlanta (humid)      25–35%      yes
Houston (very humid) 35–50%      critical
Miami (tropical)     40–50%      critical

In Phoenix, sensible-only sizing is close enough. In Houston, it undersizes by 35–50%.

The practical shortcut

If you know the entering and leaving wet-bulb temperatures, you can look up enthalpy from psychrometric tables or a calculator:

Entering: 95°F db / 75°F wb → h = 38.5 BTU/lb
Leaving:  55°F db / 54°F wb → h = 22.7 BTU/lb

Total load = 4.5 × CFM × (38.5 − 22.7)

No need to separate sensible and latent. Enthalpy captures both in one subtraction.

For quick psychrometric property lookups including enthalpy from any two known air properties, use the psychrometric calculator on CalcEngineer.

Quick Tip: Heat Pump Tonnage — Why Cooling Capacity Does Not Equal Heating Capacity

Evgenii Konkin — Thu, 24 Sep 2026 03:30:05 +0000

A heat pump is an AC that reverses direction. Same compressor, same refrigerant. But the tonnage it delivers in heating mode is not the same as in cooling mode.

Why capacity drops in heating

In cooling mode, the system rejects heat to outdoor air at 95°F. The temperature lift is moderate.

In heating mode, the system extracts heat from outdoor air at 30°F. The temperature lift is larger and there is less heat available to extract.

Mode       Outdoor    Indoor    Lift     Capacity
──────────────────────────────────────────────────
Cooling    95°F       75°F      20°F     3.0 tons (rated)
Heating    47°F       70°F      23°F     2.8 tons
Heating    30°F       70°F      40°F     2.0 tons
Heating    10°F       70°F      60°F     1.5 tons

The same 3-ton unit delivers only 2 tons at 30°F and 1.5 tons at 10°F.

The sizing dilemma

Mild climate (Houston, Atlanta):
  Cooling load: 3.5 tons
  Heating load: 2.0 tons
  → size for cooling (3.5T), heating is covered

Cold climate (Minneapolis, Chicago):
  Cooling load: 2.5 tons
  Heating load: 4.0 tons
  → size for heating, but then oversized for cooling

Moderate climate (Nashville, Charlotte):
  Cooling load: 3.0 tons
  Heating load: 3.0 tons
  → depends on specific house and insulation

Four options for cold climates

Option                           Pros                    Cons
────────────────────────────────────────────────────────────────
1. Oversize for heating          covers cold nights       short cycles in summer
                                                         poor humidity control

2. Auxiliary heat strips         simple, no extra unit    expensive to operate
                                                         $0.15-0.25/kWh electric

3. Dual-fuel (HP + gas furnace)  efficient above 30°F    two systems to maintain
                                 gas below 30°F          higher install cost

4. Cold-climate heat pump        rated to -15°F          higher equipment cost
                                 no backup needed        newer technology

The balance point

The outdoor temperature where heat pump capacity equals building heat loss:

Above balance point: heat pump covers the load alone
Below balance point: supplemental heat required

Typical balance points:
  Standard heat pump:        30–35°F
  Cold-climate heat pump:    5–15°F

If your winter design temperature is below the balance point, the heat pump alone cannot maintain setpoint on the coldest nights.

The quick check

1. Calculate cooling tonnage (area × climate factor / 12,000)
2. Calculate heating load (BTU/hr from Manual J or simplified method)
3. Check heat pump capacity at winter design temperature
4. If heating load > HP capacity at design temp → add supplemental heat

The cooling calculation is the starting point, but heating is what determines whether the system needs backup.

Heating Season Delta-T: The Ranges That Change When You Switch From Cooling to Heating

Evgenii Konkin — Wed, 23 Sep 2026 03:26:34 +0000

Delta-T in heating mode uses the same formula but different expected ranges.

ΔT = T_supply − T_return

Note: in heating mode, supply is warmer than return (reversed from cooling).

Heating delta-T ranges by equipment type

Equipment              ΔT (°F)     ΔT (°C)
───────────────────────────────────────────
Gas furnace            40–70       22–39
Electric furnace       40–60       22–33
Heat pump (heating)    25–35       14–19
Hot water coil         20–40       11–22

Compare to cooling:

Residential DX coil    14–22       8–12
Chilled water          10–12       5.5–6.7

Heating delta-T is 2–3x higher than cooling for the same equipment because combustion and electric resistance generate much higher temperatures than refrigeration.

Gas furnace: high delta-T is dangerous

In cooling, high delta-T means restricted airflow — uncomfortable but safe.

In heating on a gas furnace, high delta-T means the heat exchanger is overheating:

ΔT < 40°F  →  low output (gas pressure, ignition, flame issue)
ΔT 40–70°F →  normal range
ΔT > 70°F  →  DANGER: heat exchanger overheat
                → limit switch trips
                → if limit bypassed: exchanger cracks
                → cracked exchanger = CO in supply air

Most common cause of high delta-T in heating:

1. Dirty filter (restricts airflow)
2. Collapsed return duct
3. Undersized return grille
4. Blower at wrong speed setting
5. Failed blower capacitor

A dirty filter from last winter is the #1 pre-season find.

Heat pump: low delta-T is normal

Gas furnace supply air:   130–150°F
Heat pump supply air:      95–105°F
Body temperature:           98.6°F

Heat pump air feels "cool" compared to a furnace because the delta-T is only 25–35°F vs 40–70°F. The system compensates by running longer:

Gas furnace:  6–8 hrs/day in cold weather
Heat pump:    12–16 hrs/day in cold weather

Total daily heat delivered is similar. Delivery rate per hour is lower.

If heat pump ΔT drops below 20°F:

→ low refrigerant charge
→ defrost cycle running too frequently
→ auxiliary heat strip not energizing

Heating capacity from delta-T

Same formula as cooling:

Q (BTU/hr) = 1.08 × CFM × ΔT

Gas furnace at 1,200 CFM, 55°F delta-T:

Q = 1.08 × 1,200 × 55
Q = 71,280 BTU/hr

If nameplate says 80,000 BTU output:

71,280 / 80,000 = 89% of rated capacity → normal (duct losses)

At 35°F delta-T (something wrong):

Q = 1.08 × 1,200 × 35
Q = 45,360 BTU/hr = 57% of rated → investigate

Pre-season check

1. Replace air filter
2. Run furnace for 15+ minutes (steady state)
3. Measure supply temperature 6 ft from plenum
4. Measure return temperature before filter
5. Subtract: ΔT = supply − return
6. Compare to range for your equipment type
7. If outside range → schedule service before first cold night

Quick Tip: Why a 500 kVA Transformer Cannot Power 500 kW of Equipment

Evgenii Konkin — Tue, 22 Sep 2026 04:40:38 +0000

A transformer nameplate says 500 kVA. The building loads add up to 480 kW. Looks like 20 kW of headroom. Breakers trip anyway.

The conversion

kW = kVA × Power Factor

A 500 kVA transformer at different power factors:

Power Factor    Available kW    Typical loads
─────────────────────────────────────────────────
1.00            500 kW          resistive heaters only
0.95            475 kW          LED lighting, electronics
0.90            450 kW          mixed commercial
0.85            425 kW          motors + mixed loads
0.80            400 kW          motor-heavy industrial

At PF 0.85 (typical commercial building):

Available = 500 × 0.85 = 425 kW
Demand = 480 kW
Deficit = 55 kW → overcurrent trip

Why this catches people

The nameplate shows one number: 500 kVA. Most people read it as "500 kW capacity." That is only true if every load is purely resistive with PF = 1.0.

Real buildings have motors (compressors, pumps, fans, elevators), VFDs, and electronic equipment. Combined PF typically runs 0.80–0.90.

Building type        Typical PF    500 kVA delivers
────────────────────────────────────────────────────
Office (modern)      0.90–0.95     450–475 kW
Retail               0.85–0.90     425–450 kW
Hospital             0.80–0.85     400–425 kW
Industrial           0.75–0.85     375–425 kW
Data center          0.95–0.99     475–495 kW

Data centers are high PF because server power supplies include PFC circuits. Everything else is lower.

Sizing the transformer correctly

If building demand is 480 kW at PF 0.85:

Required kVA = kW / PF
Required kVA = 480 / 0.85
Required kVA = 565 kVA

Select a 750 kVA transformer (next standard size above 565). Not 500.

Standard transformer sizes (kVA):
75, 112.5, 150, 225, 300, 500, 750, 1000, 1500, 2000, 2500

The derating stack

Transformers lose additional capacity from:

Factor                 Typical loss
──────────────────────────────────────
Power factor < 1.0     5–25%
Ambient temp > 40°C    reduce per ANSI
Harmonic loads (VFDs)  10–15% K-factor
Altitude > 1000m       derate per IEEE

A 500 kVA transformer in a hot mechanical room with VFD loads and PF 0.85 might effectively deliver 350 kW. That is 30% less than the nameplate suggests.

The quick check

1. Sum all load kW
2. Estimate combined PF (0.85 default for mixed commercial)
3. Divide kW by PF to get required kVA
4. Select transformer ≥ required kVA (next standard size up)
5. Check for harmonic, temperature, and altitude derating

For quick kVA to kW conversion at any power factor, use the kVA to kW calculator on CalcEngineer.

Duct Noise Is Set at the Sizing Stage, Not the Complaint Stage

Evgenii Konkin — Mon, 21 Sep 2026 03:01:19 +0000

Duct noise is proportional to velocity. Doubling velocity adds roughly 16 dB. The duct size you choose locks in the noise level permanently.

Velocity drives noise

Velocity = CFM / Duct Area

1,200 CFM through two different ducts:

18" round:  Area = 1.77 ft²   V = 679 FPM    → quiet
14" round:  Area = 1.07 ft²   V = 1,122 FPM  → noisy (+8 dB)

Same airflow. 65% more velocity. 8 dB more noise.

ASHRAE recommended max velocities

Space type           Max FPM    NC rating
──────────────────────────────────────────
Residential          400–700    NC-25
Libraries, studios   500–700    NC-25–30
Private offices      600–800    NC-30–35
Open offices         700–1,000  NC-35–40
Retail, restaurants  800–1,200  NC-40–45
Industrial           1,000–2,500 NC-50+

NC = Noise Criteria. Lower NC = quieter space = lower max velocity.

Where noise gets worse: fittings

Straight duct noise is predictable. Fittings generate turbulence noise 10–20 dB louder:

Fitting                    Added noise
──────────────────────────────────────
Sharp 90° elbow (no vanes)  +12–18 dB
Vaned 90° elbow             +4–8 dB
Tee branch                  +8–14 dB
Abrupt transition           +6–12 dB
Takeoff (no scoop)          +8–15 dB

A vaned elbow is 8–12 dB quieter than unvaned at the same velocity. That is the difference between hearing the duct and not hearing it.

Rectangular duct adds breakout noise

Round duct:        curved surface resists vibration
                   minimal panel noise

Rectangular duct:  flat panels vibrate (drumming)
                   wider panel = lower frequency = louder
                   36" wide panel >> 12" wide panel

This is another reason extreme aspect ratios fail — a 36×6 duct is not just higher friction, it is also louder than a 12×8 duct of the same area.

The cost comparison

Fix at design stage:
  Select 18" instead of 14" duct
  Cost difference: ~$2–5/linear foot
  For 100 ft run: $200–500

Fix after construction:
  Duct silencer: $2,000–10,000 installed
  Duct replacement: $5,000–25,000
  Ceiling repair: additional cost
  Tenant disruption: priceless

The check

Before committing to a duct size:

1. Calculate velocity: V = CFM / Area
2. Look up NC rating for the space type
3. Compare velocity to ASHRAE max for that NC
4. If over → select next larger duct size

This takes 30 seconds. The noise complaint lasts the life of the building.

Phantom Loads: How to Find the Equipment That Burns Money 8,760 Hours a Year

Evgenii Konkin — Fri, 18 Sep 2026 03:59:15 +0000

Phantom loads are equipment that draws power 24/7 even when the building is unoccupied or the device is nominally off.

Common phantom loads in commercial buildings

Equipment                    Standby Power    Qty    Total
──────────────────────────────────────────────────────────
Computers (sleep mode)       3–10 W each      200    600–2,000 W
Monitors (standby)           1–5 W each       200    200–1,000 W
Vending machines             300–700 W each   5      1,500–3,500 W
Copiers (ready mode)         20–100 W each    10     200–1,000 W
Network switches/APs         10–50 W each     30     300–1,500 W
──────────────────────────────────────────────────────────
Typical office total                                 3–9 kW continuous

Why phantom loads are expensive

The energy formula:

kWh = kW × hours

A 5 kW phantom load running every hour of the year:

5 kW × 8,760 hrs = 43,800 kWh
43,800 × $0.15 = $6,570/year

The 66% problem

Hours in a year:          8,760
Typical occupied hours:   3,000 (50 hrs/wk × 50 wks + overtime)
Unoccupied hours:         5,760 (66% of the year)

Operating loads run during 3,000 hours. Phantom loads run during all 8,760. That means phantom loads accumulate 2.9× more hours than occupied-only equipment.

The night audit

Walk the building at 10pm on a weeknight when it should be empty:

1. Note every screen that glows
2. Note every light that is on
3. Note every device that hums
4. Note every indicator LED

Each one is drawing power that nobody budgeted for.

The meter check

Compare overnight base load to daytime peak:

Overnight / Peak = base load ratio

< 25%:  excellent phantom load control
25–35%: typical, acceptable
35–50%: high, investigate
> 50%:  excessive, immediate action needed

Cost per device category

Device              Watts    Hours/yr    kWh/yr    Cost/yr
───────────────────────────────────────────────────────────
1 computer (sleep)  8 W      5,760       46        $6.91
200 computers       1,600 W  5,760       9,216     $1,382
1 vending machine   500 W    8,760       4,380     $657
5 vending machines  2,500 W  8,760       21,900    $3,285
10 copiers (ready)  500 W    5,760       2,880     $432

Note: computers and copiers use unoccupied hours (5,760). Vending machines run all 8,760.

The fix

Most phantom loads are eliminated through operations, not capital:

Fix                              Savings estimate
─────────────────────────────────────────────────────
Computer auto-shutdown policy    $1,000–$2,000/yr (200 PCs)
Scheduled outlet strips          $500–$1,500/yr
Vending machine occupancy ctrl   $2,000–$3,000/yr (5 units)
After-hours lighting sweeps      $1,000–$3,000/yr
Network equipment scheduling     $500–$1,000/yr

Total potential: $5,000–$10,000/yr for a mid-size office building. Implementation cost: near zero for policy changes, $500–$2,000 for smart strips and timers.

Quick Tip: Solar Gain Through Windows — The Cooling Load That Changes by the Hour

Evgenii Konkin — Thu, 17 Sep 2026 03:33:39 +0000

Solar gain is the most variable component of cooling load. It changes with orientation, time of day, season, and glass type.

The formula

Q_solar = Window Area × SHGC × Incident Solar Radiation

Where:

SHGC = Solar Heat Gain Coefficient (0 to 1)
       Modern low-e glass:   0.22–0.40
       Older single-pane:    0.65–0.85

Orientation makes a 7x difference

Same 100 sq ft window, same building, same day (July, 4pm):

Orientation    Solar Gain (BTU/hr)    Why
──────────────────────────────────────────────────────
West           15,000–20,000          direct sun at hottest hour
South          8,000–12,000           sun high, partially blocked by overhang
East           3,000–5,000            direct sun was morning (already passed)
North          2,000–3,000            diffuse sky radiation only

West is worst because peak solar and peak outdoor temperature hit at the same time.

The afternoon spike

8am–2pm:  building runs fine
3pm:      west offices start overheating
4pm:      supply air cannot keep up
5pm:      complaints, thermostat wars

Cause: solar gain was averaged in the calculation
       system sized for noon average, not 4pm peak

SHGC comparison

For a 200 sq ft west-facing window at 4pm peak (100 BTU/hr/ft² incident):

Old single-pane (SHGC 0.75):
  Q = 200 × 0.75 × 100 = 15,000 BTU/hr

Modern low-e (SHGC 0.25):
  Q = 200 × 0.25 × 100 = 5,000 BTU/hr

Difference: 10,000 BTU/hr = 0.83 tons

Replacing the glass cuts solar gain by 67% on that wall.

Shading effectiveness

External shading (overhang, fin):     blocks 50–90% of direct sun
                                      most effective on south walls
                                      ineffective on west (low sun angle)

Internal shading (blinds):            blocks 10–30% of heat gain
                                      solar energy already inside
                                      blind absorbs and re-radiates as heat

External shading before the glass > internal shading after the glass. Always.

The quick check

If solar gain > 20% of total cooling load:
  → size system for 3pm–5pm peak, not noon average
  → check west-facing glazing separately
  → consider SHGC upgrade or external shading

Why Gas Furnaces Lose 4 Percent Capacity Per 1,000 Feet of Elevation

Evgenii Konkin — Wed, 16 Sep 2026 03:07:52 +0000

Gas-fired equipment is rated at sea level. At altitude, it delivers less heat. The reason is air density.

The derating rule

Capacity at altitude = Rated capacity × (1 − 0.04 × altitude in thousands of feet)

Or equivalently:

Derating factor = 1 − (0.04 × h/1000)

Where h is elevation in feet above sea level.

Example: 100,000 BTU furnace

Altitude    Factor    Actual Output
──────────────────────────────────────
Sea level   1.00      100,000 BTU/hr
2,000 ft    0.92       92,000 BTU/hr
3,000 ft    0.88       88,000 BTU/hr
5,000 ft    0.80       80,000 BTU/hr
7,500 ft    0.70       70,000 BTU/hr

A 100,000 BTU furnace in Denver (5,280 ft) delivers ~80,000 BTU.

Why it happens

Air at altitude has the same oxygen percentage (21%) but lower density:

Sea level: ρ = 1.225 kg/m³
5,000 ft:  ρ = 1.005 kg/m³ (−18%)

Each cubic foot of air contains 18% fewer oxygen molecules. The burner gets less oxygen per cycle. Less oxygen = less combustion = less heat.

The density ratio:

ρ = P / (R × T)

P at 5,000 ft ≈ 84,556 Pa (vs 101,325 at sea level)
Ratio = 84,556 / 101,325 = 0.834

The 4% per 1,000 ft rule approximates this as 0.80 at 5,000 ft — slightly conservative, which is appropriate for safety-critical combustion equipment.

The sizing mistake

Heating load calculation shows 90,000 BTU/hr needed.

At sea level:

Select 100,000 BTU furnace
Margin: 100,000 − 90,000 = 10,000 BTU (11%)
Result: adequate

At 5,000 ft (same furnace):

Actual output: 100,000 × 0.80 = 80,000 BTU
Deficit: 90,000 − 80,000 = 10,000 BTU short
Result: house cannot hold setpoint on coldest nights

Correct sizing at 5,000 ft:

Required input = 90,000 / 0.80 = 112,500 BTU
Select: 120,000 BTU furnace
Actual output: 120,000 × 0.80 = 96,000 BTU
Margin: 96,000 − 90,000 = 6,000 BTU (7%)

Combustion air also needs correction

The burner needs more volume of air to get the same oxygen:

Sea level: 15 ft³ of air per ft³ of natural gas
5,000 ft:  15 / 0.80 = 18.75 ft³ of air per ft³ of gas

Combustion air openings must be 20% larger at 5,000 ft. Undersized openings cause:

Incomplete combustion
Carbon monoxide production
Sooting and heat exchanger fouling
Premature equipment failure

Not all equipment derates the same

Atmospheric burners:     full 4%/1000 ft derating applies
Induced draft:           full derating applies
Power burners:           may partially compensate
Sealed combustion:       may auto-adjust fuel-air ratio
Condensing (modulating): check manufacturer specs

Always check the installation manual for altitude-specific requirements. Some manufacturers offer high-altitude orifice kits instead of applying blanket derating.

For quick air density checks at any altitude and temperature, use the air density calculator on CalcEngineer.

Quick Tip: How to Find the Condensation Plane Inside a Wall

Evgenii Konkin — Tue, 15 Sep 2026 03:24:38 +0000

Condensation forms where the temperature drops below the dew point. On a pipe, that is visible. Inside a wall, it is hidden.

The temperature gradient

In winter, each layer of a wall is at a different temperature:

Indoor air:    70°F
Drywall:       69°F
Insulation:    70°F → 25°F (across R-15)
Sheathing:     25°F
Air gap:       22°F
Cladding:      20°F
Outdoor air:   20°F

The temperature at each interface:

T_interface = T_indoor − (ΔT_total × R_cumulative / R_total)

Where:

ΔT_total = T_indoor − T_outdoor = 50°F
R_total  = 20

At the sheathing inner face:

R_cumulative = R_drywall(0.5) + R_cavity(15) + R_studs(2.5) = 18
T = 70 − (50 × 18/20) = 70 − 45 = 25°F

The dew point check

Indoor air at 70°F and 40% RH:

Dew point ≈ 45°F

Compare to each interface:

Interface              Temp    vs Dew Point   Risk
──────────────────────────────────────────────────
Drywall surface        69°F    > 45°F         safe
Insulation cavity      varies  crosses 45°F   ⚠️
Sheathing inner face   25°F    < 45°F         CONDENSATION
Sheathing outer face   22°F    < 45°F         CONDENSATION

The condensation plane is at the sheathing — 20°F below the dew point.

Why this destroys buildings

Visible pipe condensation:
  → noticed in days
  → fixed with insulation
  → cost: $50

Hidden wall condensation:
  → noticed in months or years
  → mold, rot, structural damage
  → cost: $5,000–$50,000+

The vapor barrier rule

Cold climate:     vapor barrier on INTERIOR (warm side)
Hot-humid climate: vapor barrier on EXTERIOR (warm side)
Mixed climate:    smart vapor retarder (adjusts with humidity)

Wrong side = moisture trapped against cold surface = guaranteed failure

The quick check

1. Calculate indoor dew point from T and RH
2. Calculate temperature at each wall interface (R-value ratio)
3. If any interface < dew point → condensation risk
4. Fix: add exterior insulation, lower indoor RH, or add vapor retarder

Why Climate Zone Changes AC Tonnage More Than Square Footage

Evgenii Konkin — Mon, 14 Sep 2026 03:11:03 +0000

AC tonnage is not set by square footage. It is set by how much heat the system has to remove per hour. And that depends on where the building is.

The tonnage formula

Cooling Load = Envelope + Occupants + Equipment + Lighting
Tonnage = Cooling Load (BTU/hr) / 12,000

The envelope load dominates in most buildings:

Envelope = Area × Climate Factor × (Ceiling Height / 8)

Climate factors by region

Climate          BTU/ft²    Example cities
──────────────────────────────────────────────
Cool             20–25      Seattle, Portland, Minneapolis
Moderate         25–30      Denver, Chicago, DC
Hot-humid        30–35      Houston, Dallas, Atlanta, Miami
Hot-dry          35–40      Phoenix, Las Vegas, Tucson

These factors account for the design outdoor temperature, solar gain patterns, and humidity levels typical for each region.

Same house, four cities

A 2,000 sq ft house with 9 ft ceilings, 4 occupants, 500W equipment, 200W lighting:

City        Climate    BTU/ft²   Envelope    Internal   Total      Tons
──────────────────────────────────────────────────────────────────────────
Seattle     Cool       22        49,500      2,400      51,900     4.3
Denver      Moderate   27        60,750      2,400      63,150     5.3
Dallas      Hot        32        72,000      2,400      74,400     6.2
Phoenix     Hot-dry    38        85,500      2,400      87,900     7.3

Same house. Tonnage nearly doubles from Seattle to Phoenix.

Why "400 sq ft per ton" fails

The most common rule of thumb:

Tonnage = Area / 400 to 600
2,000 sq ft / 400 = 5.0 tons

This gives 5 tons everywhere.

Seattle: needs 4.3T, gets 5.0T → oversized by 16%
  → short cycling, poor humidity control, wasted energy

Phoenix: needs 7.3T, gets 5.0T → undersized by 32%
  → runs continuously, never reaches setpoint on peak days

The rule was calibrated for a moderate climate. It fails everywhere else.

Ceiling height multiplier

The formula uses volume, not area:

8 ft ceilings:  2,000 × 8 = 16,000 ft³
10 ft ceilings: 2,000 × 10 = 20,000 ft³
12 ft ceilings: 2,000 × 12 = 24,000 ft³

The ceiling height adjustment:

Factor = actual height / 8

8 ft:  1.0×  (baseline)
9 ft:  1.125×
10 ft: 1.25×
12 ft: 1.5×

A house with 12 ft ceilings in Phoenix needs 50% more tonnage than the same footprint with 8 ft ceilings.

When internal gains change the picture

In Phoenix, envelope is 85–90% of total load. Internal gains barely matter.

In Seattle, envelope is 60–70%. Internal gains are 30–40%.

Seattle office with heavy equipment:
  Envelope:  49,500 BTU/hr (70%)
  Internals: 21,000 BTU/hr (30%)
  Total:     70,500 BTU/hr = 5.9 tons

Seattle bedroom (same area):
  Envelope:  49,500 BTU/hr (93%)
  Internals:  3,600 BTU/hr (7%)
  Total:     53,100 BTU/hr = 4.4 tons

Same area. Different internal loads. 1.5 tons difference. In hot climates, this gap shrinks because the envelope load overwhelms everything else.

Low Delta-T Syndrome: The Chilled Water Problem That Wastes Pump Energy

Evgenii Konkin — Fri, 11 Sep 2026 04:00:29 +0000

Chilled water systems are designed for a specific delta-T. When that number drops, pump energy explodes.

The capacity formula

Q (BTU/hr) = 500 × GPM × ΔT

Where:

500 = constant (8.33 lb/gal × 1.0 BTU/lb·°F × 60 min/hr)
GPM = water flow rate
ΔT  = return temperature − supply temperature

Design vs actual

Design conditions:

Supply:  44°F
Return:  56°F
ΔT:      12°F
Flow:    400 GPM

Q = 500 × 400 × 12 = 2,400,000 BTU/hr = 200 tons

When delta-T drops to 6°F (common in real buildings):

Supply:  44°F
Return:  50°F
ΔT:      6°F
Flow:    400 GPM

Q = 500 × 400 × 6 = 1,200,000 BTU/hr = 100 tons

Half the capacity at the same pump energy.

The operator response

To recover 200 tons at 6°F delta-T:

GPM = Q / (500 × ΔT)
GPM = 2,400,000 / (500 × 6)
GPM = 800

Double the flow. But pump power follows the cube law:

P2 / P1 = (GPM2 / GPM1)³
P2 / P1 = (800 / 400)³ = 8

In practice, the system curve limits the actual increase, but pump power typically goes from 50 kW to 150–200 kW.

Design:  400 GPM,  50 kW pump,  200 tons
Actual:  800 GPM, 200 kW pump,  200 tons (recovered)

Extra pump energy: 150 kW × 3,000 hrs/yr = 450,000 kWh
Extra cost at $0.12/kWh: $54,000/year

Why delta-T drops

Three common causes:

1. Three-way valves bypassing water around coils
   → Bypass water mixes with return, lowers return temp
   → Fix: convert to two-way valves with VFD on pumps

2. Oversized coils at part load
   → Water passes through too fast to warm up
   → Fix: adjust control sequences, reduce flow at part load

3. Dirty coils (water-side fouling)
   → Scale or biofilm reduces heat transfer
   → Fix: clean and treat water chemistry

The 60-second diagnostic

1. Clamp temp sensor on CHW supply pipe
2. Clamp temp sensor on CHW return pipe
3. Subtract

Design ΔT:    10–12°F (5.5–6.7°C)
Problem zone:  < 8°F under normal load
Critical:      < 6°F

If delta-T is below 8°F during normal cooling load, the system has low delta-T syndrome.

Air-side delta-T for comparison

The same concept applies to air systems:

Q (BTU/hr) = 1.08 × CFM × ΔT

Normal ranges:

Residential cooling (DX): 14–22°F
Gas furnace:              40–70°F
Heat pump heating:        25–35°F

Both formulas measure the same thing: how much heat the fluid picks up (or gives up) as it passes through the heat exchanger.

For quick delta-T calculations with automatic capacity output for both air and water systems, use the HVAC delta T calculator on CalcEngineer.

Quick Tip: Round vs Rectangular Duct — When to Use Which

Evgenii Konkin — Thu, 10 Sep 2026 03:47:47 +0000

Same cross-sectional area. Different performance. The shape matters.

The perimeter problem

Shape              Area     Perimeter   vs Round
────────────────────────────────────────────────
Round (16.6")      216 in²  52.2 in     baseline
Square (14.7²)     216 in²  58.8 in     +13%
Rect (24×9)        216 in²  66.0 in     +26%
Rect (36×6)        216 in²  84.0 in     +61%

More perimeter = more friction, more noise, more leakage.

Equivalent diameter

Rectangular ducts have higher friction than round for the same area. Equivalent diameter tells you what round duct the rectangle behaves like:

De = 1.3 × (W × H)^0.625 / (W + H)^0.25

For 216 in² area:

Round 16.6":   De = 16.6 in (actual diameter)
Rect 24×9:     De = 15.6 in (acts 6% smaller)
Rect 36×6:     De = 13.8 in (acts 17% smaller)

The 36×6 duct has the same area as the round but behaves like a duct 2.8 inches smaller.

When to use round

✓ Ceiling space allows it
✓ Noise-sensitive spaces (studios, hospitals, offices)
✓ Long straight runs (friction compounds over distance)
✓ Exposed ductwork (aesthetics)
✓ Spiral duct available (lower cost per foot)

When to use rectangular

✓ Ceiling height limited (9" plenum vs 16" round)
✓ Routing around beams and pipes
✓ Aspect ratio stays ≤ 4:1
✓ Multiple branches in tight spaces

The rule

Aspect ratio ≤ 4:1  → acceptable trade-off
Aspect ratio > 4:1  → friction penalty exceeds space savings
Aspect ratio > 6:1  → avoid (noise, fabrication cost, energy waste)

Noise comparison

Round duct:       curved surface resists vibration
                  minimal panel drumming
                  preferred for NC 25-35 spaces

Rectangular duct: flat panels vibrate (oil-canning)
                  low-frequency rumble
                  needs heavier gauge or external bracing
                  for noise-sensitive applications