API TR 2567:2005 pdf download

API TR 2567:2005 pdf download.Evaporative Loss from Storage Tank Floating Roof Landings.
Figure 7 shows that the stock vapor concentration rose to just under 6 percent during the relatively calm period of the first 18 hours. As soon as the wind speed rose above about 4 miles per hour, however, the stock vapors rere completely dissipated. When the wind speed subsequently subsided, an increase in stock vapor concentration was again measured. The next rise in the wind speed eliminated any sustained return of stock vapors (the narrow spikes at about the 40Lh and 64ht hours were dismissed in the Trinity report as resulting from fluctuations in the instrument readings that occur when the stock vapors are at a non-measurable level). A pattern of stock vapor dissipation in response to elevated wind speed is apparent.
3.2.2.1 Confidence in the Wind Effects Model. The equation used to model wind effects should be viewed as a placeholder, rather than as a derived characterization of the actual relationship between ambient wind speed and standing idle losses. While standing idle losses from an external floating-roof tank with a liquid heel appear to be driven by wind, the rate at which these losses occur at a given wind speed is not known. That is, the proposed equation holds a place in the model for addressing the observed phenomenon of wind effects, but there arc no data available a this time for characterizing certain variables.
In the absence of data. a placeholder was selected that is relatively simple, with the loss rate expressed as a constant times the tank diameter. While additional data may eventually justify adding terms to the proposed equation, there is no benefit to be gained by adding complexity to the equation without data to demonstrate an improvement in accuracy. The value used for the constant in this placeholder was derived from default values that were selected to generate an estimate that exceeds a rational lower bound. It would not be appropriate, then, to override the default value for one of these terms without data to guide corresponding substitutions for the other terms.
3.2.2.2 Derivation of the Wind Effects Model. The placeholder value in the TGB interim report for the rate of wind-driven standing idle loss was developed in the context of rational upper and lower hounds. The lower bound recognizes that the wind-driven estimate of emissions should never be less than the emissions estimated when wind is neglected.
3.2.3 Internal or External Floating-Roof Tanks That Drain Dry.
Standing idle losses from either an internal or an external floating-roof tank that has been drained dry are modeled as the evaporation of a thin layer of liquid clinging to the bottom of the tank. The absence of a liquid heel implies that the tank bottom is designed to allow withdrawal of virtually all free flowing liquid (i.e., drain-dry tanks). When a drain-dry tank has been completely emptied, the only stock liquid available to evaporate is that remaining on wetted surfaces of the tank interior. This evaporation of liquid clinging to a surface is termed clingage loss.
The 1998-1999 field testing demonstrated that the limited quantity of stock liquid that remains in drain-dry tanks is insufficient to sustain daily replenishment of stock vapors tinder the floating roof. This has already been observed in Figure 7, where stock vapors do not return after the wind has flushed out those stock vapors that were generated initially. Figure 8 presents the trends in nominal saturation level for all four test tanks, where it is apparent that the two drain-dry tanks (Test Tanks I and 4) do not sustain their initial level of saturation, whereas the saturation level does remain fairly constant in the tanks with a liquid heel (Test Tanks 2 and 3). It appears, then, that neither of the daily standing idle loss mechanisms (i.e., thermal breathing and wind effects) would apply to a drain-dry tank. The standing idle loss mechanism for drain-dry tanks is the evaporation of the thin layer of liquid clinging to the bottom of the tank. This would he a one-time event, rather than a daily event, and thus the estimate of standing idle loss for a drain-dry tank is independent of the number of days that the tank stands idle.