ASHRAE LV-11-015-2011 Analysis Tools and Guidance Documents for Evaluating and Reducing Vulnerability of Buildings to Airborne Threats-Part 1 Literature Review.pdf

《ASHRAE LV-11-015-2011 Analysis Tools and Guidance Documents for Evaluating and Reducing Vulnerability of Buildings to Airborne Threats-Part 1 Literature Review.pdf》由会员分享，可在线阅读，更多相关《ASHRAE LV-11-015-2011 Analysis Tools and Guidance Documents for Evaluating and Reducing Vulnerability of Buildings to Airborne Threats-Part 1 Literature Review.pdf（9页珍藏版）》请在麦多课文档分享上搜索。

1、2011 ASHRAE 817ABSTRACT This paper classifies and describes analysis methods,tools, and simulation programs that allow prediction ofairborne chemical/biological dispersal and transport dynam-ics in indoor environments subject to different risk scenarios.These are the building blocks for related anal

2、ytical treatmentof the overall problem involving risk assessment, risk manage-ment, and identifying cost-effective mitigation measures.These methods are distinguished by the level of mathematicaland scientific rigor in modeling the phenomena, in the spatialand temporal resolution in solving the mode

3、ling equations,and in the types of boundary conditions and the numericalparameters that appear in the model. The paper also describesvarious general guidance documents and vulnerability assess-ment protocols and software available in the open-source liter-ature to assess and reduce vulnerability in

4、buildings due toairborne threats and risks.INTRODUCTIONAssessing and reducing the vulnerability of buildings toairborne chemical and biological (CB) threats has been thesubject of major and numerous efforts in the last decade, andhas resulted in a large body of published knowledge. Thispaper reviews

5、 the existing open-source literature, availableprotocols, and analysis tools in order to classify them in termsof scope of applicability and state of development. The liter-ature review is not meant to be exhaustive, but every effort hasbeen made to identify representative and unique material. Thisp

6、aper starts with a classification of analysis methods, tools,and simulation programs that allow analytical prediction ofairborne chemical and biological agent dispersal and transportdynamics in indoor environments subject to different riskscenarios. This is followed by a more general screening ofdoc

7、uments, protocols, and software that practitioners can useto evaluate the vulnerability of buildings and improve thesecurity of occupants. BUILDING VULNERABILITY ANALYIS TOOLSThis section classifies and describes analysis methods,tools, and simulation programs that allow analytical predic-tion of ai

8、rborne CB dispersal and transport dynamics in indoorenvironments subject to different risk scenarios. These are thebuilding blocks for related analytical treatment of the overallproblem involving risk assessment, risk management, andidentifying cost-effective mitigation measures. These meth-ods are

9、distinguished by: (1) the level of mathematical andscientific rigor in modeling the phenomena, (2) the spatial andtemporal resolution in solving the modeling equations, (3) thetypes of boundary conditions and the numerical parametersthat appear in the model, and (4) the degree of specificity, i.e.,t

10、he tool may not have been developed for risk analysispurposes per se but could be used for vulnerability analysis ifthe simulation results are analyzed external to the tool; forexample, the CONTAM program developed by National Insti-tute of Standards and Technology (NIST) (Walton and Dols2008).Compa

11、rtmental ModelsCompartmental modeling is a special type of linearsystem modeling that is well developed and has had somesuccess in the fields of biomathematics and environmental andchemical engineering. Godfrey (1983) defines a compartmen-tal system as consisting of “a finite number of homogeneouswe

12、ll-mixed, lumped subsystems, which exchange with eachAnalysis Tools and Guidance Documents for Evaluating and Reducing Vulnerability of Buildings to Airborne ThreatsPart 1: Literature ReviewT. Agami Reddy, PhD, PE Steven SnyderFellow ASHRAE Associate Member ASHRAEJustin Bem William Bahnfleth, PhD, P

13、EStudent Member ASHRAE Fellow ASHRAET. Agami Reddy is SRP Professor of Energy and Environment in the School of Architecture and Landscape Architecture and a professor in the Schoolof Sustainability and Steven Snyder is a graduate student in the School of Architecture and Landscape Architecture, Ariz

14、ona State University,Tempe, AZ. Justin Bem is a mechanical engineer with James Posey Associates Inc., Baltimore, MD. William Bahnfleth is director of the IndoorEnvironment Center and a professor in the Department of Architectural Engineering, The Pennsylvania State University, University Park, PA.LV

15、-11-0152011. American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions, Volume 117, Part 1. For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without

16、ASHRAES prior written permission.818 ASHRAE Transactionsother and with the environment so that the quantity or concen-tration of material within each compartment may be describedby a first-order ordinary differential equation (ODE).” Thisformalism of compartmental modeling is particularly appro-pria

17、te for preliminary “what-if” studies and for calibratingmodels against limited and sparse direct concentration-measurement data (Evans 1996a, 1996b). It can provideinsights into issues of problem identification, such that thedesign of field measurements can be suitably tailored as aresult. On the ot

18、her hand, compartmental modeling provideslimited detail in terms of spatial concentration distribution.Finally, the numerical values of the model parameters appear-ing in the equations, which are to be viewed as lumped param-eters, have to be determined with care. The set of modelingequations can be

19、 compactly formulated in a state-space repre-sentation. Depending on how one solves this set of equations,one can distinguish between three different approaches appli-cable to the context of airflows in buildings.1. Closed form analytical solutions can be determined for anumber of practical one-zone

20、 and two-zone buildingscenarios. In addition to computational simplicity, theyallow convenient model parameterization, which is veryuseful for model calibration using field data, allowsuncertainty analysis to be performed easily, and providesinsight into which physical parameters need to bemeasured

21、with more accuracy (Reddy and Bahnfleth2007). 2. Spreadsheet programs can be used to simulate air andcontaminant flows by reducing the number of zones in abuilding to a small, manageable number of aggregatedwell-mixed zones. Such a representation may beadequate for assessing the vulnerability of bui

22、lding occu-pants to inhaled dose (Kowalski 2003). The rationale forthe simpler approach is that well-defined boundaryconditions (necessary for detailed engineering simula-tion) are hard, if not impossible, to characterize practi-cally, given that event scenarios have large uncertaintyand variability

23、. Further, the spreadsheet simulationapproach allows calibration to be achieved with relativeease, in addition to which parametric and sensitivity anal-yses can be performed in a straightforward manner. 3. General-purpose building simulation programs generatenumerical solutions for, in effect, as ma

24、ny zones as onewishes. The most common programs are CONTAM(Walton and Dols 2008), RISK by the EnvironmentalProtection Agency (EPA), and COMIS by LawrenceBerkeley National Laboratory (LBNL 2003). Lorenzetti(2002a) assesses several multizone software, whilecomputational aspects are reviewed in Lorenze

25、tti(2002b). Such multizone models provide informationabout the dynamic behavior of room average indoorconcentrations at a considerable reduction in computa-tional effort as compared to the detailed CFD modelingapproach. Effects such as infiltration due to stack andwind pressure differences; interzon

26、al flows throughwalls, doors, and windows; and stack effects in stairwellsand elevator shafts can be explicitly considered. Note thatin IAQ-relevant studies, it is not simply the air within thespace that is treated as a compartment, but floors, carpets,drapes, HVAC ducts, and other components with w

27、hichindoor pollutants can interact are also modeled ascompartments. Shortcomings of this approach are that itcannot model zones that are poorly mixed, nor can itcapture bidirectional and vertical stratification effects. Asa result, occupant exposure levels predicted by thesemodels are room-average v

28、alues that may be under orabove actual exposure levels. These models have beenextensively used for building-related analysis involvingboth design and performance evaluation (e.g., Musserand Persily 2002; Persily et al. 2007), as well as forassessing the casualty impact from biological agents(Kowalsk

29、i and Bahnfleth 2003). The effort involved inmodeling the building and then calibrating it againstmonitored data is time consuming, with the latter aspectstill an area of active research (e.g., Price et al. 2003 andFirrantello et al. 2005). Detailed Deterministic Simulation ProgramsReports by Stenne

30、r et al. (2001) and Sohn et al. (2004)identify current simulation models for determining dispersionand migration of CB agents within and around the exterior ofbuildings, and review the capabilities and limitations ofselected models. Sohn et al. (2004) suggested a classificationscheme of existing sim

31、ulation software, which is modifiedinto the grouping as follows:1. Computational fluid dynamics (CFD) models givedetailed, if not always accurate, predictions of the fateand transport of CB agents in HVAC systems and of thespatial concentration distribution in the room/buildingunder well-defined bou

32、ndary conditions. Severalcommercial software packages can be used; however,Airpak by ANSYS Inc., Flovent by Flometrics Inc., andFEM3MP by Lawrence Livermore National Laboratory(LLNL) are especially meant for simulating air andpollutant transport within the built environment.2. Real-time dispersion-d

33、eposition-causality models (alsoreferred to as operational models) have been developedfor use by emergency managers, battle commanders, andbuilding designers and maintenance personnel to limit theloss of life in simulating hazardous material releasescenarios over regional scales of 1 to 100 km. They

34、 linkreal-time weather data, topological data, dispersionmodeling, and population data to the resulting populationexposure level. These models are based on statisticaldispersion methods rather than on fundamental fluid flowequations, and use 10 40 m grid cell resolution. Sohn etal. (2004) list a doz

35、en such programs, and discuss a fewin some detail. 2011 ASHRAE 8193. Environmental regulatory models are also unsuited forIAQ modeling but are meant for the prediction of hazardmigration in the outdoor environment. These requireterrain depiction, modeling of vertical and horizontalturbulence, atmosp

36、heric convective mixing, and model-ing of aerodynamic effects. Sohn et al. (2004) provide alist that includes 7 preferred models, 10 simplifiedscreening models and 34 alternative models, all of whichare meant for environmental regulatory purposes.AERMOD by USEPA is considered the most accurateregula

37、tory model available. Typical spatial area suited tothese programs is 1 100 k, with 1 m numerical grid reso-lution. The bibliographic report by Chapman and Thomas(2007) also discusses a number of software tools.Probabilistic Modeling and Simulation ProgramsThe previous two groups of analysis methods

38、 of analyseswere deterministic in their approach in that all inputs to themodel were characterized by single values (presume knownwithout any uncertainty). The usual manner to study the effectof uncertainties on the system response using such tools is toperform a sensitivity analysis. This is genera

39、lly done in asomewhat ad-hoc manner, involving only parameters selectedby the user and with deviations from baseline values selectedsomewhat arbitrarily. However, there is an alternativeapproach, called probabilistic or stochastic modeling, wherethe uncertainties are treated in a more structured man

40、ner andwhich is more appropriate for risk-based modeling. Inessence, the inputs are assumed to be random variables and arerepresented by probability distributions rather than point esti-mates. Numerous sources of uncertainty surrounding the issue ofdeveloping assessment protocols for enhancing the s

41、ecurity ofbuildings vis-vis indoor-related IAQ threats have beenpointed out by Kunreuther (2002) and Reddy and Bahnfleth(2007). These include (1) uncertainties in the probability ofoccurrence of an “extreme IAQ event”; (2) in framing theboundary conditions and in the input parameter specifications;(

42、3) accuracy of the system behavior prediction; (4) of the dose-response relationship for the particular CB agent; (5) the valid-ity of the consequence function to the owner; and (6) the costuncertainty associated with implementing the necessary miti-gation measures identified. Given the above uncert

43、ainties, it isappropriate to review some probabilistic modeling softwarethat is commercially available. One can distinguish betweenthe two following types:1. General-purpose risk analysis software is available intwo platforms: add-on and stand-alone. Crystal Ball(Decisioneering Inc.) and -Risk (Pali

44、sade Corp.) areperhaps the most widely known add-on software forspreadsheet programs (which allow forecasting and opti-mization to be done along with risk analysis) and adoptMonte Carlo simulation by which to generate numeroustrials. GoldSim (2005) is an example of stand-alone soft-ware that allows

45、complex analyses and modeling to beperformed. It is a highly graphical, object-oriented simu-lation platform suitable for performing dynamic, proba-bilistic simulations to support management and decisionmaking in engineering, science, and business. Anotherversatile modeling and simulation platform f

46、or assessingrisks in macro-environmental effects (which includeambient air fate and transport models, water-borne fateand transport models, dose response models, etc.) is theFramework for Risk Analysis in Multimedia Environ-mental Systems (FRAMES), developed by Pacific North-west National Laboratory

47、 (PNNL) (Sohn et al. 2004).FRAMES is not a modeling and simulation tool as such,but a software interface structure that allows legacymodels, databases, and software from disparate sourcesto communicate in a plug-and-play manner. Not onlydoes it allow data transfer between different specializedprogra

48、ms, but it also has a suite of tools for integrating,analyzing, and visualizing data, and for performing sensi-tivity and uncertainty analyses.2. Specialized software, of which GoldSims “ContaminantTransport Module” is particularly appropriate for under-standing and predicting the migration of mass

49、(e.g.,contaminants) in environmental systems. This is anextension of GoldSim (2005), which allows convenientmodeling of the release, mass transport, and fate of mate-rials and pollutants within complex environmentalsystems. Chapman and Thomas (2007) have identified a largenumber of resources, both in the form of reports and softwaretools, on hazard identification (both natural and man-made),risk assessment, and risk management software. Some perti-nent risk reduction guidance and protocols available in theopen literature are described in the next section. FULL-RISK ASSESSM