Underfloor Air Distribution
What is it Underfloor Air Distribution?
Underfloor air distribution (UFAD) is a method of bringing conditioned air into a space through an underfloor plenum that is below a raised floor system.
Per a Trane engineering newsletter on UFAD systems from 2001 [1], the potential benefits of a UFAD system include improved thermal comfort due to local direct control of supply airflow and reduced ventilation rates because of higher zone air-distribution effectiveness due to stratification. Space stratification, or the temperature differential in a conditioned space that changes with height, and thermal performance of underfloor plenums are key drivers in load calculations for UFAD systems.
UFAD systems typically deliver conditioned air to the space through diffusers mounted in the raised floor. This swirling supply air mixes with air from within the space, creating a mixed-air region in the lower portion of the space. Then, temperature stratification occurs in the upper portion of the space (see figure below).
Displacement ventilation (DV) systems are similar to UFAD systems; however, it typically involves delivering slow-moving air into the space from diffusers placed low in sidewalls [2]. In this system, heat sources (such as people or equipment) induce the local airflow from the floor toward the ceiling. The UFAD diffuser airflows are typically larger than for DV systems, which reduce the ceiling height necessary for acceptable temperature stratification. Additionally, displacement ventilation does not include any effect of heat transfer into the underfloor plenum on the cooling airflow.
Though in real life while many conventional air distribution systems experience some stratification, in TRACE™ 700 overhead distribution systems assume well mixed spaces where UFAD and DV systems have stratification. TRACE™ 700 has been able to run UFAD and displacement ventilation calculations since version 6.2 in December 2008.
How does stratification and underfloor plenum heat transfer affect space conditions?
Space stratification is a simple concept where several layers of air at a range of heights have different temperatures. Stratification is important to consider when low velocity air distribution or distribution at the floor level is being considered.
In the case of the UFAD system, temperature stratification in the conditioned space will result in higher temperatures at the ceiling (higher than the room thermostat set point) which will change the heat transfer dynamics in a room and between floors.
To compare between cooling airflow quantities used by a UFAD system and an overhead system, there is a defined acceptable comfort condition for standing occupants in a stratified room. These requirements were taken from the ASHRAE Journal in October 2007on calculations for UFAD by Bauman [3].
- Average occupied zone temperature is calculated as the average of measured temperature profile from the ankle level (4 inches) to head level (67 inches), is equal to the desired thermostat set point.
- The occupied zone temperature difference that is calculated from the head to ankle does not exceed the maximum limit of 5 degrees F as specified by ANSI/ASHRAE Standard 55-2010.
UFAD systems provide supply air at higher temperatures when compared to overhead systems based upon the maximum difference specified by ASHRAE 55-2010. As a result of the higher supply air temperature, the potential exists to increase airside economizer operation hours in some climates.
The figure below shows a schematic diagram of an example room air temperature profile identifying the important factors of stratification.The temperature differentials come from the EnergyPlus room air model method UCSD UFAD Interior Underfloor Air Distribution which is also used as the basis for the TRACE™ 700 room air temperature profiles.
T67 = room air temperature at a height of 67 inches (i.e., head height)
T4 = room air temperature at a height of 4 inches (i.e., ankle height)
Ts = underfloor supply air delivered by diffuser
Tset = thermostat set point at 48 inches above floor
Toz,avg = average room air temperature in occupied portion of space
ΔToz = temperature difference between the head and the ankle of a standing person in the occupied zone
Furthermore, the radiant contribution of the room loads can have a major effect on temperatures in the room and plenums. Including the radiative heat transfer from the ceiling to the floor and the conduction through the raised floor into the supply plenum, there will be a 35% to 45% heat gain into this supply underfloor plenum and a net heat loss from the return plenum of 10% to 15%[3].
As cool supply air passes through the underfloor plenum, it is exposed to heat gain from both the floor slab below and the raised floor above. This heat gain can be large and result in a temperature gain to the supply air inside the plenum. This will then lead to a possible increase in supply airflow to counter the heat gain.
UFAD Other Considerations
Supply air leakage from the underfloor plenum into the room can often be in the range of 10% to 30% of the design airflow depending on whether or not the underfloor is pressurized [3]. This leakage is accounted for in the design calculations by including the estimated maximum air leakage rate into the predicted airflows. Any leakage above 20% may influence the amount of stratification in the room as well as the ability for the room to control the thermostat set point. However, the leakage will change the load in the space and offset the required amount of increased supply airflow to the diffusers.
How is space stratification calculated in TRACE™ 700 with underfloor plenum heat transfer?
The UFAD algorithmin TRACE™ 700 is derived from research by the Center for the Built Environment (CBE) at the University of California as well as some methodology in the EnergyPlus program [2].
The key aspectto determining space stratification is the temperatures from ankle height (4 inches) to head height (67 inches). These temperatures are used to determine a dimensionless temperature, Phi, which is a ratio of the set height temperature to the supply and return air temperatures.
Φ = (T- Ts)/(Tr - Ts)
T = temperature at given height above floor
Ts = supply temperature
Tr = return air temperature
This Phi value is predicted for underfloor air distribution systems as a function of Gamma – a non-dimensional parameter [4, 5, 6].According to Lin and Linden[7], the primary parameters for controlling stratificationcan be defined in Gammaas the buoyancy flux of the heat source and the momentum flux of the cooling jets. These fluxes are derived through plume theory for the heat source and a fountain model for the diffuser flow. These are non-dimensional values. A large Gamma reduces stratification where a small Gamma produces the opposite effect.The theoretical and experimentally tested model has been implemented in EnergyPlus by using a slightly different definition of Gamma.
Gamma is used in two different calculations, for interior zones as well as perimeter zones as defined by Liu [8]. According Perimeter zones account for the line plumes generated by heat gains from exterior windows and walls. However, TRACE™ 700 will only use the interior zone relationship in the stratification calculation.
The interior zone relationship is as follows:
Γ = Non-dimensional parameter representing the ratio of buoyancy to inertia forces in the zone
Q = room airflow [m3/s]
Ad = Diffuser effective area
Cos θ = discharge angle for diffuser flow [m2]
n = number of diffusers
m = number of plumes (i.e., occupants or workstations)
W = room cooling load [kW]
The perimeter zone relationship is as follows:
Γ = Non-dimensional parameter representing the ratio of buoyancy to inertia forces in the zone
Q = total perimeter zone airflow [m3/s]
Ad = Diffuser effective area [m2]
Cos θ = cosine of discharge angle for diffuser flow
n = number of diffusers;
WL = zone extraction rate per unit length of zone [kW/m], the zone cooling load (supply and return plenum cooling loads are not included) divided by the length of the external wall of the perimeter zone considered
In addition, the temperatures can drastically change based on the type of diffuser used to distribute the air. The choice of a diffuser will give its effective area and a discharged angle factor which will be used in the Gamma equation listed above. The diffuser effective area and angle factor were determined through laboratory testing and manufacturer’s catalogues.
Diffuser Type / Supply Fan or System Type / Zone Type / Diffuser Effective Area, ft2(m2) / Angle Specific to Diffuser Type, θ [°] / Angle Factor Specific to Diffuser Type, cos θSwirl / Constant Volume / Interior / 17.05 (0.011) / 28 / 0.883
Variable Area / Variable Volume / Interior and perimeter / 54.25 (0.035) / 45 / 0.707
Linear Bar Grill / Constant Volume / Perimeter / 38.75 (0.025) / 15 / 0.966
Horizontal Swirl / Displacement Vent / Interior and perimeter / 9.3 (0.0060) / 73 / 0.292
Each diffuser has a Gamma-Phi relationship that determines the stratification within the room. This relationship was tested experimentally throughCBE [2]and forms the basis for the UFAD stratification algorithm in TRACE™ 700.However, when selecting a variable air volume UFAD system in TRACE™ 700, the diffuser type is Variable Area. When selecting a constant volume UFAD system, the diffuser type is Swirl. All displacement ventilation systems assume a diffuser type of horizontal swirl.
Zone / Diffuser Type / Gamma / Phi 4in / Phi 67inInterior / Swirl / Γ < 4.3
4.3 < Γ < 46
Γ > 46 / 0.4212
0.2684 Γ0.3089
0.875 / 0.944
Interior / Variable Area / 0.745 / 0.956
Perimeter / Linear Bar Grill / Γ < 11
11 < Γ < 29.56
Γ > 29.56 / 0.47
0.2199+0.0224 Γ
0.882 / 0.882
Perimeter / Variable Area / 0.631 / 0.874
Furthermore, after determining the Phi value based on Gamma, the temperatures at head height and ankle height can be calculatedbut any downstream air leakage associated with the UFAD systemwill affect how these temperatures are calculated.Thesetemperatures will then be used in solving for the average occupied zone temperature.
The average occupied zone temperature is important as it is the acceptable comfort condition for standing occupants. Depending on the stratification in the space, theaverage Tozis typically lower than the single point thermostat temperature at 48 inches for overhead distribution systems. From an application standpoint, according to Bauman [3], to maintain equivalent comfort conditions (well-mixed to a stratified environment)it is recommended to raise the thermostat set point at 48 inches by one degree F for a ΔToz at three degrees F and by 0.5 degrees F for a ΔToz at two degrees.
In TRACE™ 700, the UFAD stratification calculations are iterative due to the constantly changing fraction of total space sensible loads being transmitted to the underfloor. The conduction calculationderived from Bauman, H. Jin, T. Webster [9] is very complex; however, it is more or less a steady-state energy balance iteration that is performed until the conduction load through the raised floor and average underfloor supply air temperature come into balance.
FSLO = Φoz(1 - FSLU)
FSLO = Fraction of Total Space Sensible Loads to Occupied Layer
FSLU = Fraction of Total Space Sensible Loads to Underfloor
Once the space sensible load to the underfloor is determined based on conduction through the underfloor over the overall total space sensible load, the fraction is used in conjunction with the dimensionless temperature of the average occupied zone to determine a fraction of total space sensible loads to the occupied layer.
FSUS = (1 – FSLO – FSLU)
FSLO = Fraction of Total Space Sensible Loads to Occupied Layer
FSLU = Fraction of Total Space Sensible Loads to Underfloor
FSUS = Fraction of Total Space Sensible Loads to Upper Stratified Layer
Every time the fraction of total space sensible loads to the occupied layer changes, the required airflow to the space changes, which also changes Gamma and Phi. The change in the required space airflow will also affect the convective and radiant exchange in the underfloor supply plenum.The conductive loss will then increase or decrease through the raised floor which will impact the space sensible load fraction to the underfloor.
During both the Design and System Simulations the program uses an algorithm to check whether hourly conditions will actually produce temperature stratification for a particular room. For example, if the supply air is warmer than the room air, the driving force for the convective plumes will not be present. Similarly, if the room is not occupied and has low internal heat gains, the driving force that causes the heat to rise towards the ceiling exhaust will not affect the flow pattern. Chilled ceiling panels or passive chilled beams that supplement the cooling capacity of displacement ventilation or UFAD systems will create downdrafts in the occupied portion of the space, thus reducing the amount of convective fraction that goes to the upper stratified layer. For any hour that these conditions occur the program assumes that stratification does not occur and the associated room air is fully mixed.
Furthermore, stratification will not occur in TRACE™ 700 under these conditions:
1)If UFAD or DV system is NOT selected
2)Supply air temperature is warmer than the room air temperature
3)Room has low internal heat gains, i.e., (InternalSpaceSensibleNoStratification - ConductionLossToUnderfloorPlenum)/RoomFloorArea < 0.1 Btuh/sqft)
4)Chilled ceiling panels or passive chilled beams are operating at more than 5% of their aux cooling capacity
5)Kill stratification during heating or deadband mode (otherwise this will increase heatingload by moving cooling space load to the upper return air layer)
What are the key input parameters for UFAD systems in TRACE™ 700?
There are several parameters in TRACE™ 700 that are important to the methodology behind the UFAD system load calculations.
Supply air path / duct location
For Supply air path / duct location, there are four options to choose from. Additionally, these options will also be affected bysupply duct / other losses if entered.
- No duct, Supply via Underfloor Plenum
- Ducted through Underfloor Plenum
- Supply via Sidewall Displacement Ventilation
- Supply via Underfloor Displacement Ventilation
If No duct, Supply via Underfloor Plenum or Supply via Underfloor Displacement Ventilationis selected, a complex heat exchange can occur between the slab, raised floor panels, side walls (if they exist), and the supply air as it flows through the underfloor air plenum. If the room is warmer than the underfloor plenum, the supply air traveling in that plenum will pick up heat from the room. If the floor slab has absorbed any heat or is exposed to outdoor conditions or an unconditioned space, then the supply air temperature will be impacted based on conduction heat gain/loss into the underfloor plenum.
If Ducted through Underfloor Plenum or Supply via Sidewall Displacement Ventilation is selected, then no heat exchanged is assumed between the underfloor plenum and the supply air. The Ducted through Underfloor Plenum is also used in configurations where directed airflows are used to move the air through an underfloor plenum with very minimal interaction with the floor slab and raised floor.
As discussed earlier, there can be significant leakage in UFAD systems due to the floor panel configuration as well as poor construction/sealing. In general, downstream leakage is assumed to leak into the occupied layer above the current space’s underfloor plenum while the upstream leakage is assumed to bleed into the corridor return air path, thus the plenum temperature will not be affected but the return air condition to the AHU coils will be impacted.
The upstream leakage fraction is the upstream leakage flow divided by the supply fan design airflow and it is a constant for all part load ratios. The downstream leakage fraction is a constant fraction of the mixing box airflow, which may vary during system operation for a VAV. The sum of the two fractions at design will not equal the total leakage fraction.
The auxiliary cooling coil losses to plenum %, the auxiliary cooling coil devices which are either recessed or suspended from the ceiling (such as passive, active chilled beams and chilled ceiling panels) will lose a portion of their capacity in the form of radiation and/or convection to the ceiling plenum. This loss will be proportional to the hourly cooling load.
For UFAD systems, in the Advanced options of Create Systems, If the Supply Duct / Air Path Location = Return Air, the supply ducts are assumed to be located in the ceiling plenum and so both the upstream and downstream duct leakage is assumed to leak into the associated return plenum, either raising or lowering the plenum temperature accordingly; however, if no ceiling plenum exists for a particular room, then the leakage is assumed to bleed into the “corridor” return air path and so will still effect the return air condition seen by the AHU coils. If the return air path is Room Direct, all leakage occurs directly to the room instead of the ceiling plenum.
If the Supply Duct / Air Path Location = Other, then the leakage is assumed to bleed to ambient and the supply air lost and its air conditioning effect wasted.
For all other Supply Duct / Air Path Location locations, i.e., supply air delivered via an underfloor and/or displacement ventilation, the site of the downstream vs upstream leakage differs. In order to distinguish where the leaks occur, the downstream leakage is assumed to leak into the occupied layer above the current room’s underfloor plenum while the upstream leakage is assumed to bleed into the “corridor” return air path and so will not affect the plenum temperature but will still affect the return air condition seen by the AHU coils.
To better explain upstream/downstream leakage, let’s take a look at a basic example ofa system with 10% upstream leakage and 10% downstream leakage for a standard VAV system. If the supply fan airflow is 10,000 CFM, then the upstream leakage is 1,000 CFM (10% of 10,000 CFM) and 9,000 CFM reaches the terminal mixing boxes. The downstream leakage is therefore 900 CFM (10% of 9000 CFM) and 8,100 CFM reaches the room via the supply diffusers. This means that a total of 1,900 CFM or 19% of the 10,000 CFM supply fan airflow has leaked from the ducts.