A comprehensive modal emissions model for light-duty cars and trucks isbeing developed under the sponsorship of NCHRP Project 25-11. Modeldevelopment has been described previously for vehicles operating underhot-stabilized conditions. A modal emissions model is presented for vehi-cles operated under incremental soak-time conditions. The Federal TestProcedure (FTP) measures vehicle emissions after a 24-h soak time dur-ing Bag 1 testing and vehicle emissions after a 10-min soak time duringBag 3 testing. Vehicle incremental soak-time emissions refer to vehicleemissions after intermediate variable soak times of between 10 min and24 h. Recent research shows that most on-road vehicles experience soaktimes of between 10 min and 24 h during daily driving; thus, there is strongdesire to model vehicle emissions under such circumstances. An interme-diate soak emission model has been developed on the basis of second-by-second emissions measurements generated at the College of Engi-neering Center for Environmental Research and Technology, Universityof California–Riverside vehicle testing facility by using the FTP Bag 1and Bag 3 test cycles. The modeling results are based on a composite vehicle concept in which more than 300 tested vehicles are compositedinto two dozen vehicle technology groups. The modeling approach is afuel-based physical modal emissions model in which vehicles’ fuel use,engine-out emissions, air/fuel equivalence ratio, catalyst efficiencies, andtailpipe emissions are modeled individually as a function of variable soaktime. Since the developed model is based on modeled vehicle fuel con-sumption under any given driving cycle, it not only is capable of predict-ing vehicle emissions under variable soak time for any given test cycle butalso is capable of predicting emissions under different starting test cycles.The College of Engineering Center for Environmental Research andTechnology (CE-CERT) at the University of California–Riversideis developing a comprehensive modal emissions model for light-duty vehicles under sponsorship of NCHRP [NCHRP Project 25-11,(1–5)]. The overall objective of this project is to develop and vali-date a modal emissions model that accurately reflects the impacts ofspeed, engine load, and start conditions on exhaust emissions undera comprehensive variety of driving characteristics and vehicle tech-nologies. In this project, approximately 300 in-use vehicles have beenrecruited and tested on a single-roll 1.22-m (48-in.) dynamometerover three different driving cycles: (a) the Federal Test Procedure[FTP (6)], (b) the high-speed US06 cycle (7), and (c) a speciallydesigned modal emission cycle [MEC01, which is described else-where (1)]. For each of these cycles, second-by-second engine-outand tailpipe carbon dioxide (CO2), oxygen (O2), carbon monoxide(CO), oxides of nitrogen (NOx), and hydrocarbon (HC) emissions dataare being collected. On the basis of these measured emissions data, amodal emissions model is being developed to estimate vehicleemissions under several operating modes:(a) The cold start period, that is, the first few minutes after thevehicle is started;(b) Stoichiometric operation,the predominant mode of operationwhen the vehicle’s air/fuel ratio is at the proper stoichiometric ratio;(c) Enrichment events, that is, when excessive load conditionsare placed on the engine (e.g., during sharp accelerations and steepgrades) and the air/fuel ratio is commanded rich; and(d) Enleanment events, which typically occur with sharp decel-eration or load reduction events, during which time the air/fuel ratiois lean and incomplete combustion or misfire occurs.Details of this model development have been described elsewhere(1–5); however, this paper focuses on modeling of emissions underincremental soak-time conditions and what impacts different cold-starttest cycles have on the catalyst light-off time and overall emissions.FTP Bag 1 measures vehicle emissions after a 24-h or longer soak time, and FTP Bag 3 measures vehicle emissions after a 10-minsoak time. Recent research shows that most on-road vehicles experi-ence soak times of between 10-min and 24-h during daily driving(8–10). The Environmental Protection Agency (EPA) has examinedthe causes of postsoak emissions with data from the EPA Soak/StartTest program and a preliminary program called the Albany CooldownStudy that gathered real-world engine and catalyst cooldown profiles.The data from these programs indicated that increased emissions fol-lowing intermediate soaks arise in three ways: rapid catalyst cool-down following key off, slow catalyst thermal recovery following arestart, and manufacturer calibration strategies in response to the start-up condition. In the most updated EMFAC7G model, the CaliforniaAir Resources Board (CARB) has included vehicle emissions undervariable soak time (11). Thus, there is a strong interest from regula-tory agencies and environmental groups to fully understand vehicleemissions under variable intermediate soak times.EPA has also developed a new Soak Control Cycle (SC01) that isto be used to control emissions after intermediate soaks. Initial idlesand start driving are addressed in SC01 by incorporating the EPAStart Cycle (ST01) in its entirety. The balance of SC01 is composedof two microtrips of moderate driving selected from the in-use sur-vey database. EPA considers the SC01 cycle preferable because ithas speeds and power levels that are more representative of in-usestart driving behavior (7). CARB has also introduced a differentcold-start cycle, the LA92 Bag 1 cycle, to simulate vehicle cold-startdriving under a contemporary California driving environment (12).Whether these new starting cycles can or should be eventuallyincluded in legislation is beyond the scope of this paper. Neverthe-less, it shows that there is a demand for modeling of vehicle cold-start emissions under variable starting cycles.In this paper, preliminary analysis and modeling of vehicle emis-sions under cold-start conditions are presented. In the past, EPA andSystems Applications International have developed a fuel-based cold-start emissions model on the basis of extensive regression analysis(13,14). The present analysis and modeling are based on measureddata from CE-CERT’s NCHRP testing program. To assist with modeldevelopment, the emissions traces of composite vehicles are used. TheModal-Based Intermediate Soak-TimeEmissions ModelingFENG AN, MATTHEW BARTH, GEORGE SCORA, AND MARC