My familiarity is primarily with fluorescent dye tracers and fluorometer, mostly in streams and to help trace extent of pesticide contamination from aerial spraying.  Therefore my comments are field based and may not address your specific problem.  With lab soil column many replications may be done to help determine how much to add and moisture level of soils to obtain desired results.  So under those conditions, reducing tracers may still produce results, and you may help define how much is appropriate for your type study.

Sometimes too much is added to insure results in the field testing, rather than injecting tracer with no results.  How much to add may be based on experience from past work.  Sometimes constant injection is used to assess tracer concentration and find peak to estimate various things like dilution, loading, time of travel, etc. With a tracer, once added in field site, the site is basically contaminated if there is no timely result from adding small amount, adding more later to the existing site adds uncertainty if something shows up, was it from first application or second as there is high variability of hydraulic conductivity and whether an appropriate flow path was identified between injection and measurement site.  One may not always predict vectors of groundwater flow path and dispersion spread from the surface features in the field, and this could change not only with geology, topography, soil or substrate differences, but also somewhat with level of water table.  A certain amount of tracer may be retained, mixed in soil water, adsorbed or absorbed in the transference, another consideration.

Dear Ali,

As Mr Hansen thoroughly explained, the main thing for any transport experiment is to get an appreciable level of concentration in the effluent. Besides for the specific case of nano-particle (NP) transport experiments you may consider some other points as follows.

I think one reason for multi-pore-volume injecting is representing the field condition in remediation projects where large amounts of NP are need to remove the contaminants. Besides, in most of column transport studies in addition to the breakthrough curve (BTC), retention profile (RP) is also measured which is substantially affected by the way of injecting the NP dispersion. If it is a case, you may need to inject limited number of pore volumes because if you inject continuously, you might not be able to separate the aqueous phase from the retained phase in the sectioned porous media at the end of the experiment.

Moreover, the most important mechanisms of NP transport, namely, deposition, release, straining, aggregation, blocking, ripening, and size exclusion, influence or are affected by the load of particles, i.e., the injection concentration of NPs and the number of pore volumes (PVs). For instance, if blocking is the dominant phenomena, once all the porous media surfaces are covered by NPs, the rest of injecting NPs would pass through the packed column. However, if ripening is the dominant mechanism, more deposition of the particles brings about further retention onto previously deposited ones. Thus, it is beneficial to have an imagination of the expected regime of transport. This mainly depends on the solution chemistry (ionic strength (IS) and pH), flow regime, and particle characteristics. I might say roughly, if the IS is low and the pH is high, you can get results with small number of PVs injection. One way of rough pre-estimation is to use colloid filtration theory relationships to find the approximate expected C/C0 for the given length etc. To see if straining might dominate or not, you may use the available criteria of the ratio of particle size to grain diameter; >0.0017, straining might be significant. 

Hope this helps

When a short pulse is applied, Breakthrough curve can hardly reach C/C0=1 for a tracer or a NP solution, beacause of the dispersivity of the porous medium. Cmax/C0 depends also on the length of your experimental colums.

In general, in presence of NP retention (blocking), if you increase the injected NP volume, you increase Cmax/C0.

Thank you all for cooperation and sharing valuable ideas. I would also like to share the below paper which concludes short pulse would have its own advantages and then discuss more..

http://onlinelibrary.wiley.com/doi/10.1029/97WR00298/pdf