Although the use of spray ice as a construction material in the Arctic has gained wide acceptance, little quantitative information has been published on the time dependent deformations experienced by spray ice structures due to internal and externally-applied loads. This paper describes a finite element analysis of the time dependent, viscoelastic (creep) behavior of Amoco's MARS spray ice island. Two dimensional, plane strain elements were utilized to model the island. Emphasis was on the deformation of the structure due to its self-weight and the weight of the drilling rig, as well as from lateral displacements occurring in the surrounding ice sheet. The analysis was limited to secondary (steady state) creep behavior. Required values for all physical parameters were obtained from actual island measurements, including those from in situ pressuremeter tests carried out at the island.
The analytical results were compared to actual island performance characteristics, measured during its operating life performance characteristics, measured during its operating life over the winter and spring of 1985–86. The results show reasonably good agreement between modeled behavior and actual island performance. Observed discrepancies are discussed in the paper. paper.
Introduction
The use of spray ice as a construction material for offshore exploration islands in the Arctic has rewarded operators with tremendous environmental and economic advantages over the more traditional approaches used previously. In the Beaufort Sea, spray ice islands have become a fairly routine option for prospects located in favorable water depths and ice regimes. The prospects located in favorable water depths and ice regimes. The controlling issue then becomes whether or not adequate drilling time can be provided following completion of island construction.
The first of these grounded islands to be operationally as an exploratory drilling platform was the MARS ice island, constructed by Amoco Production Company. MARS was sited some three miles offshore from Cape Halkett in the northwestern corner of Harrison Bay (see Figure 1). Built during the winter of 1985–86 in a water depth of approximately 26 ft (8 m), MARS confirmed the concept feasibility previously demonstrated by the Sohio Test island of the previous winter.
A number of other grounded spray ice structures have been constructed in both the U.S. and Canadian Beaufort Seas. Extension of this technology beyond the limiting confines of the landfast ice zone has also been investigated. These efforts, in conjunction with earlier construction of floating spray ice structures in the high Arctic, have led to a great deal of confidence in the suitability of spray ice as a construction material.
Knowledge of the mechanical properties of spray ice, however, is generally limited to specifics obtained in conjunction with the referenced projects and associated testing programs. No generalized information, based on comprehensive laboratory testing of the material in all its forms and at all temperatures, pressures, strain rates and time frames of engineering interest, is pressures, strain rates and time frames of engineering interest, is currently available. Additionally, the particular application for which the material is intended (e.g., whether for a protective term or for an exploration island) in large measure dictates the construction-technique utilized and the nature of the resultant material. This has particular relevance with respect to the creep behavior of the material created and its importance to the serviceability of the spray ice structure. Thus, a protective berm around a structure may be rapidly constructed with no particular concern about excessive settlements during its lifetime, while an exploration island must be carefully built up with due regard to the sensitivity of a drilling operation to both total and differential settlements.
The organization of this paper will be according to the following format. First, a brief review of the available literature relative to the creep behavior of spray ice will be provided. This will he followed by a discussion of some specifics relative to the finite element modelling of the MARS island.
P. 297
