Key takeaways
SHR is a capacity split, not an extra safety factor: (SHR = Q_s/Q_t).
In high-sensible IT rooms, SHR mismatch can create a sensible shortfall even when “total tons” look adequate.
The fastest check is: (Q_{s,avail} = SHRcdot Q_t) and (Q_{t,needed} = Q_s/SHR).
Airflow and coil conditions shift SHR; treat SHR as “at rated conditions”, not a constant.
What this guide covers (and what it doesn’t)
Covers: neutral formulas, what to request from submittals, and sample calculations (illustrative).
Doesn’t cover: full psychrometric modeling, detailed coil selection, or vendor-specific performance claims.
Assumptions: sea level. All example numbers are explicitly labeled.
Key Takeaway: In server room sizing, “tons” without SHR is incomplete. Always convert total capacity into usable sensible capacity at the stated conditions.
Step 1 — Define your terms and units
Inputs
IT sensible load (Q_s) (kW)
Candidate unit total capacity (Q_t) (tons or Btu/h)
Candidate unit SHR at your expected operating point
Formulas
SHR definition (standard HVAC definition):
[ SHR = frac{Q_s}{Q_t} ]
Total capacity split:
[ Q_t = Q_s + Q_l ]
Sensible/latent from SHR:
[ Q_s = SHRcdot Q_tquadquad Q_l = (1-SHR)cdot Q_t ]
Done when… you can distinguish sensible vs latent vs total (and you know which one your IT load is).
Reference (definition): The Engineering ToolBox provides the standard SHR definition and interpretation in its note on Sensible Heat Ratio (SHR).
Step 2 — Sensible heat ratio (SHR) server room cooling: estimate the load SHR
Action
As a first-pass assumption for an IT room with limited people load, treat the load as high sensible (often (SHRapprox 0.85) to (0.95)).
If your design includes meaningful outdoor air, infiltration, or strict humidity targets, expect more latent fraction (lower load SHR).
This is also where the precision cooling vs comfort cooling SHR difference matters: comfort systems are commonly optimized for lower SHR (more latent fraction), while precision/computer-room systems are commonly optimized for higher SHR.
Done when… you have a load SHR range you can use for mismatch checks.
Internal context: For background on why IT rooms tend to be high-sensible and why comfort vs precision systems differ, see precision vs comfort cooling for server rooms.
Step 3 — Convert IT sensible load into required total tons at a given SHR
This step answers: “If my load is mostly sensible, how many total tons do I need if the unit SHR is X?”
Action
Convert IT sensible load from kW to Btu/h: [ Q_s,(Btu/h) = Q_s,(kW)times 3412 ]
Convert sensible load into required total capacity at the unit SHR: [ Q_t = frac{Q_s}{SHR} ]
Convert Btu/h to tons: [ text{tons} = frac{Q_t,(Btu/h)}{12{,}000} ]
Sample calculation (illustrative)
Assume:
IT sensible load (Q_s = 100,kWRightarrow 341{,}200,Btu/h)
Case A: unit SHR = 0.90 [ Q_t = frac{341{,}200}{0.90} approx 379{,}100,Btu/h Rightarrow 31.6,tons ]
Case B: unit SHR = 0.70 [ Q_t = frac{341{,}200}{0.70} approx 487{,}400,Btu/h Rightarrow 40.6,tons ]
Output
A clear “required total tons vs SHR” sensitivity, even before detailed modeling.
Done when… you can show how total required tons changes when SHR changes.
Step 4 — Compute “usable sensible capacity” from a unit’s total tons (the derate check)
This step answers: “Given a unit rated at X tons and SHR=Y, what sensible kW do I actually get?”
Action
Convert total tons to total capacity in Btu/h: [ Q_t,(Btu/h) = 12{,}000times text{tons} ]
Convert total capacity into sensible capacity: [ Q_{s,avail} = SHRcdot Q_t ]
Convert Btu/h to kW: [ Q_{s,avail},(kW) = frac{Q_{s,avail},(Btu/h)}{3412} ]
Sample calculation (illustrative)
Assume a unit is rated:
(Q_t = 30,tonsRightarrow 360{,}000,Btu/h)
If SHR = 0.90:
(Q_{s,avail}approx 0.90times 360{,}000 = 324{,}000,Btu/hRightarrow 95,kW)
If SHR = 0.70:
(Q_{s,avail}approx 0.70times 360{,}000 = 252{,}000,Btu/hRightarrow 74,kW)
Done when… you can spot a “30-ton unit” that is actually a ~74 kW sensible unit at the stated SHR.
Step 5 — Tie SHR to airflow: CFM per ton, ΔT, and why assumptions matter
A common airside sensible approximation at standard conditions is:
[ Q_s = 1.08cdot CFMcdot Delta T ]
Where (Delta T) is return-to-supply dry-bulb difference (°F).
Action
If you know (Q_s) and a design (Delta T), estimate the airflow you need:
[ CFM = frac{Q_s}{1.08cdot Delta T} ]
Sample calculation (illustrative)
Assume:
(Q_s = 100,kWRightarrow 341{,}200,Btu/h)
(Delta T = 20°F)
[ CFM approx frac{341{,}200}{1.08times 20} approx 15{,}800,CFM ]
Why this affects SHR
Changes in airflow and coil operating point can shift how much of the coil’s work is sensible vs latent, changing SHR.
Done when… you can back-calculate a CFM requirement from sensible load and see whether a submittal’s airflow is plausible.
Reference (airside equations): Belimo’s HVAC formulas and calculations guide summarizes common field equations such as (1.08cdot CFMcdot Delta T).
Step 6 — Coil impacts: ADP, bypass factor, and when SHR “moves” lower than expected
If you need a practical explanation of why SHR changes with conditions, these coil concepts are enough for early design checks:
Entering air dew point: indicates moisture in the air.
Apparatus Dew Point (ADP): effective coil surface dew point temperature.
Bypass factor (BF): fraction of air not fully contacting the coil.
A common approximation:
[ BF = frac{T_{out}-T_{ADP}}{T_{in}-T_{ADP}} ]
How to interpret this in sizing terms
Lower effective coil temperature (lower ADP) and/or better contact (lower BF) generally increases condensation → more latent fraction → lower SHR.
In high-sensible IT rooms, excessive latent fraction can drive avoidable humidity swings or control energy (depending on your humidity strategy).
Done when… you can explain to a reviewer why “SHR at AHRI rating” might not match “SHR at our return air and airflow.”
Step 7 — Submittal-sheet checklist (what to request, what to calculate)
Use this checklist to make SHR math actionable:
Get the rating conditions
entering/return dry-bulb and humidity (or dew point)
leaving/supply condition
airflow (CFM)
Get the capacities at those conditions
total capacity (Q_t)
SHR (or sensible capacity directly)
Run two calculations
Usable sensible: (Q_{s,avail} = SHRcdot Q_t)
Total needed: (Q_{t,needed} = Q_s/SHR)
Compare to your load case
If (Q_{s,avail}) is below IT sensible load, you have a sensible shortfall regardless of “tons.”
Internal resources (optional)
System taxonomy: precision air conditioning options for server rooms
Low-friction next step
If you want a reusable worksheet, request a “SHR-based sizing checklist” that includes the input fields above and the formulas from Steps 3–5, so your team can repeat the same checks across vendor submittals.
