Optimal ventilation and filtering strategies Barcelona - May, 7 2015 - PowerPoint PPT Presentation

Optimal ventilation and filtering strategies Barcelona - May, 7 2015 for air quality in carriages P . Blondeau, M. O. Abadie University of La Rochelle, France Open Day on Air quality in rail subway systems, CSIC

Expansion of the existing rail public transport network of Paris 2 4 new lines and 2 expanded lines :  + 70 km of high speed metro Underground  + 30 km of metro  +150 km of light rail 72 stations, 57 new Frame of the study

Motivations of SGP for IAQ studies 3 Construction and operation based on sustainable development principles French national plan on IAQ (2013): « Action S: undertake actions as a way to improve IAQ in subway systems »  Setup of mandatory monitoring of IAQ in subways C outdoor < C station < C platform < C carriage (except Taipei) Objectives 500 Underground train 1 Identify air quality main issues  1212 Underground platform 450 Station and key parameters for IAQ 400 Outdoor 350 Literature review: Frame of the study 300 PM 10 (µg/m³) 250  Pollutant concentrations 200 150 in metro areas 100 50 0 Target contaminants: PM 10 , PM 2.5 , NO 2 , benzene  Technical data : ventilation (stations/tunnels/carriages), tyre/steel wheels, braking system, platform equipments (e.g. window screens ), … 2 Define design guidelines for acceptable air quality in the underground indoor spaces from airflow and air quality simulations

4 Modelling strategy AIRFLOW AIRFLOW Pressures on platform System 1: window screens System 2 : TUNNEL TUNNEL / STATION Airflow rates between the Airflows in tunnels and compartments of the station ventilation sinks Frame of the study IAQ IAQ Pollutant concentrations at platfom screen doors System 4 : System 3 : STATION TUNNEL Pollutant concentrations in tunnels IAQ System 5: CARRIAGES C out

« Fresh » air Exhaust air Single zone model accounting for: 5 AHU Ventilation (including filters) C out Deposition onto surfaces Recycled air - Influence of passengers  Volume reduction  Particle emission Return air Supply air  Particle deposition on body C ind Deposition  Particle resuspension due Models and parameters to movements (body, floor) Internal Deposition emissions  Particle sink by inhaled air Resuspension 2 simulation tools : Inhaled air Design tool (steady-state) Exposure assesment (dynamic) MP 05 train carriage: V = 73 m 3 , Max. load: 162 persons  V body = 11 m 3 (1)  S body = 340 m 2 (1) Resuspension  PM10 emission rate = 100 mg/h (2) (1) H = 1.65 m, W = 65 kg, winter clothes (2) 1.6 10 6 to 3 10 6 part/h/person, density =1000 kg/m 3  Inhaled air flow rate = 68 m 3 /h (3) (3) 0.42 m 3 /h/person

Filter efficiencies for representative subway aerosol 6 Ventilation filters are classified into 16 classes and 9 classes by international standards Protection of HVAC IAQ improvement components Models and parameters 9 M M Filter class is determined based on the mean efficiency which is measured in standard test conditions using a specific aerosol  Not representative of the actual efficiency

100 Computation of the initial efficiency 7 90 of filters as a function particle size MERV 4 (G2) 80 MERV 5 (G3) MERV 6 (G3) Modified particle 70 MERV 7 (G4) size distribution 60 MERV 8 (G4) MERV 9 (M5) 50 MERV 10 (M5) MERV 11 (M6) Fresh air filter 40 MERV 12 (M6) MERV 13 (F7) 30 MERV 14 (F8) MERV 15 (F9) 20 MERV 16 (F9) Exp 10 0 Models and parameters Supply or return 0.010 0.100 1.000 10.000 Filtering efficiency as a function of particle size air filter 100% 94% Fresh air filter 93% PM10 91% 88% 90% 87% 86% 83% G2 (MERV 4) 82% Particle size distribution of a typical 80% 74% 72% subway aerosol (Midander et al, 2014) G4 (MERV 8) 71% 70% 64% 63% 60% 58% 56% 10000 52% 50% 1000 46% dN/dlogDp (particles/cm 3 ) 45% 42% 100 40% 37% 10 33% 30% 27% 1 23% 0.1 20% 16% 12% 0.01 9% 10% 0.001 1 10 100 1000 10000 100000 0% Dp (nm) G2 (MERV 4) G3 (MERV 5) G3 (MERV 6) G4 (MERV 7) G4 (MERV 8) M5 (MERV 9) M5 (MERV 10) M6 (MERV 11) M6 (MERV 12) F7 (MERV 13) F8 (MERV 14) F9 (MERV 15) F9 (MERV 16) Supply / return air filter

1 Deposition velocities on surfaces 8 Ceiling Walls Floor 0.1 Deposition velocity (cm/s) Deposition velocity of particles Friction velocity = 3 cm/s as a function of size (computed 0.01 from Lai and Nazaroff model) 0.001 0.0001 Deposition velocity of PM10 (m/h) as a function of fresh air filter 0.00001 Models and parameters 0.000001 10 0.001 0.010 0.100 1.000 10.000 Particle diameter (µm) 1 Ceiling Walls Floor 0.1 0.01 0.001 G2 (MERV 4) G3 (MERV 5) G3 (MERV 6) G4 (MERV 7) G4 (MERV 8) M5 (MERV 9) M5 (MERV 10) M6 (MERV 11) M6 (MERV 12) F7 (MERV 13) F8 (MERV 14) F9 (MERV 15) F9 (MERV 16)

Train MP05, 1 st configuration Fresh air Cooling 9 filter coil Cooled air mechanical ventilation: no recycled air, 11.5 kW cooling coil Supply air - AHU filter Simulation conditions:  162 passengers (full load)  Tunnel concentrations: 200 m g/m 3 PM 10 , 75 m g/m 3 PM 2.5 1815 m 3 /h 1815 m 3 /h Applications : design tool 220 Fresh air filter Outdoor PM 10 200 MERV 4 (G2) Indoor concentration ( m g/m 3 ) 180 MERV 5 (G3) 160 MERV 6 (G3) 140 MERV 7 (G4) WHO 120 guidelines MERV 8 (G4) 100 (2005) 80 T echnical Supply air 60 13 24-hour mean specifications 40 Internal emissions 52 Paris Annual mean 20 (passengers) 0 Supply air filter

Simulation conditions: 10 Train MP05, 1 st configuration  162 passengers (full load) Cooled air mechanical ventilation:  Tunnel concentrations: no recycled air, 11.5 kW cooling coil 200 m g/m 3 PM 10 , 75 m g/m 3 PM 2.5 100 Fresh air filter PM 2.5 90 MERV 4 (G2) Indoor concentration ( m g/m 3 ) 80 Outdoor MERV 5 (G3) Applications : design tool 70 MERV 6 (G3) WHO 60 MERV 7 (G4) guidelines 50 MERV 8 (G4) (2005) 40 30 24-hour mean 20 Annual mean 10 0 Supply air filter  Small influence of the fresh air filter on PM 2.5 concentrations  Significant efficiency of filters from MERV 11 (M6)  MERV 14 (F8) needed to meet the WHO guideline for 24-hour mean

Fresh air 2100 m 3 /h Train MP05, 2 nd configuration Return 11 filter AHU air filter Air conditioning (recycled air) Simulation conditions: Recycled air -  162 passengers  Tunnel concentrations: 200 m g/m 3 PM 10 , 75 m g/m 3 PM 2.5 7400 m 3 /h 5300 m 3 /h Applications : design tool 220 Fresh air filter Outdoor 200 PM 10 Indoor concentration ( m g/m 3 ) 180 MERV 4 (G2) 160 T echnical MERV 5 (G3) specifications 140 MERV 6 (G3) WHO 120 MERV 7 (G4) guidelines 100 MERV 8 (G4) (2005) 80 60 24-hour mean 40 Annual mean 20 0 Return air filter

Operating conditions as the optimum between IAQ and energy 12 PM 2.5 concentration – Fresh air filter = MERV 6 (G3) Normalized IAQ index  C C  max C  norm C C max min 0 < C norm < 1 (*) Applications : design tool Normalized energy index Fan power (W):      q P P  v ducts filter E  Fan energy – Fresh air filter = MERV 6 (G3) fan    P P   init . final (Pa) P filter 2  E E  min E  norm E E max min 0 (*) < E norm < 1 (*) best performance

13 Performance index C  0 < I <  (best performance) norm I E norm Optimum Fresh air filter = MERV 6 (G3) Performance index Applications : design tool 4.5 PM 2.5 4.0 Optimum 3.5 3.0 2.5 2.0 1.5 1.0 Technical specifications MERV 16 (F9) 0.5 MERV 14 (F8) 0.0 3000 MERV 13 (F7) 1500 4000 2000 5000 2500 6000 MERV 11 (M6) 3000 7000 3500 8000 4000 Fan efficiency :  fan = 0.54

Paris Train : MP05 14 220 140 PM 10 Tunnel : 200 m g/m 3 Applications : dynamic tool 200 Passengers 120 180 Number of passengers 160 100 On ground Concentration ( m g/m 3 ) section Cooling with fresh air only 140 Fresh Air : MERV6 (G3) 80 Supply : No filter Mean : 100 m g/m 3 120 100 60 80 45 m g/m 3 : Mean 2013 A/C with recycled air 40 60 Roosevelt station Fresh Air : MERV6 (G3) WHO 24-hour mean Return : MERV14 (F8) 40 20 20 WHO annual mean Mean : 30 m g/m 3 0 0 15:34:14 15:35:11 15:36:08 15:37:05 15:38:02 15:38:59 15:39:56 15:40:53 15:41:50 15:42:47 15:43:44 15:44:41 15:45:38 15:46:35 15:47:32 15:48:29 15:49:26 15:50:23 15:51:20 15:52:17 15:53:14 15:54:11 15:55:08 15:56:05 15:57:02 15:57:59 15:58:56 15:59:53

15 Easy to use models / friendly environment  System owner or subway operator Technical specifications in the frame of call for tenders  Engineers and designers Optimal ventilation / air conditioning strategies  Health authorities Assessment of exposure Possible developments  Modeling the increase of filter efficiency with particle load Model improvement  Modeling the influence of airflow rate upon filter efficiency  Improve accuracy of the energy index from information about cross sectional areas of air outlets / return duct in metro trains Conclusion  Additional criteria for optimization : filter lifetimes, replacement costs , … Acknowledgements Thanks to Klara Midander and Karine Elihn (Stockholm), Costas Sioutas and Winnie Kam (Los Angeles), Yu-Hsiang Cheng (Taipei)