Based on animal studies (mostly involving rodents), it is known that benzo[a]pyrene concentration in the air beyond certain levels increases the incidences of certain cancers in animals. However, the development of these cancers depends upon chronic exposure to higher levels of BaP. 

Dear Badria,

Cancer risk assessment of BaP (polynuclear aromatic hydrocarbons, PAHs) in ambient air is estimated by using BaP as an indicator of the mixture of carcinogenic PAHs in air and applying that to the dose–response relationship observed in epidemiologic studies. Risk assessments performed from animal experimental evaluations are also taken into consideration. Risks from inhalation exposure to BaP have been predominnatly obtained from hamster models of carcinogenesis. The World Health Organization has employed epidemiologic data gathered on coke-oven workers for risk assessment of BaP (and PAHs) in the Air Quality Guidelines for Europe. Enhanced risk for lung cancer in coke-oven workers has been utilized for quantitative risk estimates with the content of B[a]P as an indicator substance. The most plausible upper bound individual lifetime unit risk estimate associated with a continuous exposure to a fixed concentration of benzene soluble compounds of coke oven emissions in ambient air is estimated.

Reports have documented that the relationship between levels of aromatic adducts in white blood cell DNA and ambient air levels of BaP was not linear, suggetsing that other sources of BaP such as food might be critical contributors to genotoxic load.

Epidemiologic data illustrate that occupational exposure to soot, coal tar, and other PAH-containing mixtures is carcinogenic to humans. BaP in the ambient air is adsorbed onto cigarette smoke, asbestos and presumably airborne particles. These co-occurring components are also carcinogenic, and could have synergistic effects on BaP carcinogenicity. Particulates may operate through non-genotoxic (epigenetic) mechanisms at much lower concentrations present in ambient air. In such situations, the genotoxic action of BaP (and PAHs) adsorbed onto particulates could well be a more significant risk factor.

The following link might be of interest.

http://www.ncbi.nlm.nih.gov/pubmed/25345924

de la Gala Morales M, Holgado FR, Marín MR, Blázquez LC, Gil EP. Ambient air levels and health risk assessment of benzo(a)pyrene in atmospheric particulate matter samples from low-polluted areas: application of an optimized microwave extraction and HPLC-FL methodology. Environ Sci Pollut Res Int. 2015 Apr;22(7):5340-9. doi: 10.1007/s11356-014-3722-x. Epub 2014 Oct 28. PubMed PMID:  25345924. 

Thank you Dr. Kandaswami and Dr.Choubey

According to the WHO (2000) “Air Quality Guidelines for Europe (2nd Edition)”, “Based on epidemiological data from studies in coke-oven workers, a unit risk for BaP as indicator air constituent for PAHs is estimated to be 8.7x10-5 per ng/m3…. The corresponding concentrations of BaP producing excess lifetime cancer risks of 1/10000, 1/100000 and 1/1000000 are 1.2, 0.12 and 0.012 ng/m3, respectively.”

For quantitative risk assessment, following from the above researcher's report of a unit risk or slope factor of 8.7x10-5 per ng/m3, if your ambient BaP (or total PAH as BaP) concentration was 0.5 ng/m3 your cancer risk level would be approximately 1/23,000 calculated as follows:  

[ 1/(0.5 ng/m3 x 8.7x10-5 (ng/m3)-1)] = 1/22989

However, it should be noted that there are very conservative assumptions utilized in the whole risk assessment process; conditions of high exposure under occupational settings may not realistically reflect population risks from ambient exposures. Quantitative risk assessment results which assume a linear dose response would be considered worst case.

Dear Lois, you are quite right: conservative assumptions of different kind are alfa and omega of the whole risk assessment ideology and thus methodology.  It's better be safe than sorry.

USEPA IRIS-  Toxicological (CASRN 50‐32‐8) Review of Benzo[a]pyrene In Support of Summary Information on the Integrated Risk Information System (IRIS) August 2013

page 2-35, section 2.4.3

"Using linear extrapolation from the BMCL10 of 0.198 mg/m3, an inhalation unit risk of 0.5 per mg/m3, or 5 × 10-4 per μg/m3 (rounding to one significant digit), was calculated."

http://cfpub.epa.gov/si/si_public_pra_view.cfm?dirEntryID=66193