RADIOLOGICAL IMPACT OF APPLYING PHOSPHOGYPSUM TO TILLED LAND CROPPED TO ANNUAL RYEGRASS

The production of phosphoric acid by acidulation of rock phosphate wit h sulfuric acid produces phosphogypsum (PG) as a byproduct. Phosphogyp sum is primarily CaSO4 . 2H(2)O and is a potential source of S and Ca for crops. Because of the presence of small amounts of Ra-226, the U.S . Environmental Protection Agency (EPA) has imposed severe restriction s on the use of PG. The objective of the study was to evaluate the rad iological impact of applying phosphogypsum to tilled land. The experim ental land spanned over two soils (Myakka-sandy, siliceous, hypertherm ic Aeric Haplaquods and Pomona-sandy, siliceous, hyperthermic Ultic Ha plaquods) tilled and cropped to annual ryegrass (Lolium multiflorum La m.) for 3 crop-years. The PG with 792 Ra-226, 836 Pb-210, and 744 Po-2 10 Bq kg(-1) was applied all at once at the start of the study at 2 an d 4 Mg ha(-1), with no PG as control, and harrowed into the top 15-cm layer of the soil. This layer initially contained 15.9, 28.1, and 14.1 Bq kg(-1) of Ra-226, 210Pb, and Po-210 respectively. The 4 Mg PG ha(- 1) would have increased the initial soil Ra-226 and Po-210 by 9% each and Pb-210 by 5%, assuming a soil bulk density of 1500 kg m(-3). The r esults, averaged over crop-year and over the 3-crop-year period, showe d no measurable increases in Ra-226, Pb-210, or Po-210 in soil down to 90 cm sampled in layers of 15 cm, in groundwater sampled at 90 to 120 cm depth, in regrowth and mature forages, and in gamma radiation and airborne Rn-222 both measured 1 m above the plots. However, Rn-222 flu x measured at the soil surface increased by 0.74 x 10(-4) Bq m(-2) s(- 1) per Mg PG ha(-1). Thus, yearly application of 0.675 Mg ha(-1) (uppe r level of annual application rates for gypsum as S or Ca source for c rops) over a 100-year period of PG used in the study could increase Rn -222 nux from the experimental land from 8.14 x 10(-4) to 58.09 x 10(- 4) Bq m(-2) s(-1) at the end of said period, assuming no loss of Ra-22 6, the source of Rn-222 in soil and PG. This 100-year-end value, howev er, would still be less than half of 160.0 x 10(-4) Bq m(-2) s(-1), th e average Rn-222 nux for U.S. soils. However, because of the long half -life of Ra-226 (1620 years), it is important that the experimental or field rates-of-loss of PC-attributable Ra-226 in soil and soil surfac e Rn-222 flux be determined. This can be facilitated by using much hig her PG rates and more replicates than those used in the present study.