An apparatus to determine the Actual Delivered Density (ADD) of quick response and residential sprinklers under light hazard conditions was constructed and calibrated. The plume of a burning chair at different stages of fire development was simulated by the apparatus using five convective heat release rates: 110, 160, 250, 300 and 390 kW (6300, 9100, 14,200, 17,100, and 22,200 Btu/min). The ADD as a function of the convective heat release rate was determined for a residential sprinkler at ceiling heights ranging from 2.4 m to 4.6 m (8 ft to 15 ft). ADDs were obtained for single and multiple sprinklers installed using a 3.6 m x 3.6 m (12 ft x 12 ft ) spacing. The ADD of a 12.7 mm (1/2 in.) orifice, quick response sprinkler was determined for a single sprinkler directly over the ADD apparatus at a ceiling height of 3 m (10 ft). Previously reported Required Delivered Density (RDD) measurements for the reclining chair allowed suppression predictions to be made. Seven full-scale fire tests were conducted to evaluate the ADD/RDD approach to suppression prediction. Predictions were verified in five tests; suppression, however, occurred in two tests with ADD < RDD. The results of the seven tests indicate that the ADD/RDD approach used provided a conservative means for predicting suppression.
I NTRODUCTIONThe Actual Delivered Density (ADD) concept was developed at Factory Mutual Research Corporation (FMRC) as a means of predicting sprinkler performance in its Early Suppression Fast Response (ESFR) Program, The ADD of a sprinkler is &dquo;the density of water actually penetrating the fire plume and delivered onto the top of the burning array.&dquo;l In order to predict the performance of a sprinkler from ADD tests, it is necessary to compare the ADD results with the Required Delivered Density (RDD) for a particular commodity at the convective heat release rate for which first sprinkler actuation occurs. The RDD is the water application density delivered to the top surface of the burning commodity which will cause suppression. A discussion of the ADD/RDD approach to predicting suppression and its development in the ESFR Program is given in Part 1 of this study.2 In the ESFR program, a series of full-scale fire tests indicated that when the ADD provided by all the operating sprinklers was larger than the RDD value, early suppression was achieved.1 It is interesting to note that in some tests in which the commodity was between two or four sprinklers, suppression occurred even though ADD was less than RDD. Presumably the existence of other suppression mechanisms (e.g., droplet impingement on the side) not simulated by the RDD tests was responsible for this outcome.