Search

Early attempts to hydraulically fracture and connect two wells drilled at the Hot Dry Rock (HDR) site at Fenton Hill in New Mexico produced a large volume of fractured rock, but no connection. Microearthquakes triggered by fracturing indicated that the stimulated rock zones grew in unexpected directions. Consequently one of the wells was sidetracked at a depth of 2.9 km. It was redrilled into the zones of most intense microseismic activity, and flow connections were achieved. Hydraulic communication was improved by supplemental stimulation using recently developed high temperature and high pressure open hole packers. Preliminary testing indicates a reservoir with heat production capability which greatly surpasses that attained in the earlier Phase I reservoir. Longer term testing in 1987 and 1988 will provide more complete information on reservoir behaviour.

The San Juan Islands of Washington State form a geologically complex province located between the north Cascades, Vancouver Island, and the Olympic Peninsula. We have obtained 53 fission-track dates from the San Juan Islands province that help constrain its late Paleozoic to early Cenozoic tectonic and sedimentary history and its relationship to neighboring geologic terranes. The San Juan Islands can be divided into two main blocks separated by the Haro fault. South of the Haro fault, complexly deformed, metamorphosed, and probably exotic early Paleozoic to early Late Cretaceous rocks form four imbricate thrust plates separated by south- and east-dipping late Early to Late Cretaceous thrust faults. Reset zircon fission-track dates indicate that thrusting may have produced an upside-down geothermal gradient in the uppermost plate, the Decatur terrane. If present, this gradient was probably produced by conductive or frictional heating associated with a now-eroded overlying thrust fault and hot thrust plate. Cretaceous thrusting in the southern San Juan Islands was accompanied by uplift and resetting of apatite fission-track dates. In contrast to correlative rocks of the southern San Juan Islands, Upper Triassic to Lower Cretaceous rocks in and north of the Haro fault zone are essentially unmetamorphosed and only broadly folded. Apatite dates from the Upper Triassic Haro Formation and the Upper Jurassic and Lower Cretaceous Spieden Group indicate they did not participate in Late Cretaceous uplift of the southern San Juan Islands. Together with their basement (the Wrangellia terrane?), these rocks probably acted as a backstop to thrusting. The synorogenic Late Cretaceous Nanaimo basin formed north of the Haro fault in front of the advancing San Juan Islands thrust system. The age of Nanaimo deposition matches uplift (apatite) dates in the southern San Juan Islands, and detrital zircons from the Nanaimo Group yield dates consistent with southern San Juan Islands sources. Burial led to resetting of apatite dates in what is probably the deeper part of the Nanaimo basin.

Summary Phyllosilicates and zeolites grew in Columbia River sediments during hydrothermal experiments at relatively low temperatures and pressures. Although the new minerals may not be equilibrium assemblages, our results strongly support the idea that matrix minerals in graywackes may be the result of alteration of components thermodynamically unstable in the environment of diagenesis. Scanning electron micrographs show that the new minerals have formed as a mesh-like coating on original grains. The textural relationship of the new minerals to the original minerals resembles graywacke texture.

This is a letter responding to inquiries concerning the Hot Dry Rock Geothermal Energy program.

We have been asked by DOE in Washington to make program adjustments which reflect different funding scenarios for FY82. We find it impossible to plan the program until the large uncertainties regarding the cost of the Phase II fracture operations are in hand. The goal of the FY82 fracturing activities is to create "multiple" fractures, thereby connecting the two wellbores, and to flow the new system. It is our understanding that three types of downhole operations are being considered; cement plugs; perforated liners; and open hole packers.

The mean particle size of sediment carried on the bed of the Columbia River increases downstream. The Snake River and other upstream tributaries contribute mostly fine sediment carried in suspension. Many downstream tributaries produce coarser sediment, which is carried largely in the bedload. Bedload is appreciable only in the lower portions of the river, and is a small percentage of the suspended load Large migrating sand waves form the top surface of the bed. The distribution of sediment on some of the reservoir floors changes throughout the year. Sediment deposited during low discharges may be scoured from the reservoirs during periods of high discharge and transported downstream through the dams. The amount of sediment transported by the Columbia River system during a single flood may exceed that which is transported during an entire "average" year. The mineral and chemical composition of Columbia River sediment obtained from reservoir floors suggests that the sediment particles have undergone relatively little chemical weathering. Chemically unstable mineral grains and lithic fragments abound. Quartz and elements considered relatively immobile during chemical weathering show no obvious enrichment in the sediment. The ratio of quartz to all "unstable" constituents (feldspars, mafic minerals, and lithic fragments) decreases downstream, whether the latter are taken individually or as a group. Columbia River sediment is apparently derived from two principal sources: the upstream sediment from surficial deposits is derived ultimately from metamorphic, plutonic, and sedimentary rocks; and downstream sediment is derived largely from the erosion of andesitic volcanic rocks and sediments. The upstream sediment is fine grained and is transported by the Columbia in suspension. The volcanic sediment is coarser and generally forms the bedload. It is likely that these two sediment types remain more or less physically separated during transport and deposition.

Experiments with sands chemically similar to graywackes produced "matrix" minerals at 250 degrees C and a water pressure of 1 kilobar. Starting materials were moderately well sorted sands containing abundant volcanic rock fragments and no detectable clay minerals. The results support the concept that graywacke matrix may form diagenetically by hydration and recrystallization of unstable clasts. Matrix need not be a recrystallized detrial fine fraction. Texture, therefore, is not a reliable criterion for sandstone classification. For graywackes, chemical composition may be more significant.

Sediment samples from the Columbia River were analyzed for four variable sets: bulk chemistry, trace element chemistry, bulk mineralogy, and heavy mineral content. The variable sets were analyzed by discriminant function analysis and factor analysis to determine the relationships between samples in Columbia River reservoirs as reflected in their distribution in the variable space. Other statistical studies were made to determine the relationships among the variables and changes in the variables related to position in the river and grain size. The distribution of the samples in the variable space indicates changes in the sediment which can be related to changing source areas.

The mineral and chemical composition of sediments deposited in the three lowermost reservoirs of the Columbia River is remarkably similar to the composition of many graywackes. Lithic fragments are abundant. In comparison with an "average" sandstone, the sediments have low concentrations of silica and high concentrations of all other major constituents, except calcium. Sodium is more abundant than potassium. The sediments are generally better sorted than graywackes. If graywacke texture is post-depositional in origin, Columbia River-type sediments could be expected to form graywackes upon deep burial without any significant addition or removal of material.

Peridotite clasts in fanglomerate on the west margin of the Chiwaukum graben have rims of serpentine (lizardite) that formed after deposition. Interbedded sedimentary rocks are incompletely altered to laumontite and other low P-T phases. Serpentinization must have occurred under P-T conditions no higher than zeolite facies, probably less than 100°C.

Fragments of Jurassic ophiolite having U-Pb zircon ages narrowly grouped at 160 to 170 m.y. are widespread over parts of northwest Washington. The Haystack thrust fault is inferred to mark the base of the ophiolite in the San Juan Islands and adjacent Cascade foothills; other bodies of mafic and ultramafic rock in the western Cascades may be klippen of the Haystack thrust plate. The Haystack thrust fault is probably the structurally highest and possibly most extensive thrust yet recognized within a family of Late Cretaceous thrust faults in northwest Washington. The ophiolite and its time of emplacement (bracketed between about 100 and 88 m.y.) suggest a similarity with the Coast Range thrust of California which thrust Upper Jurassic ophiolite and the Great Valley sedimentary sequence over the Franciscan assemblage. However, relations in the Cascades are complicated by the extraordinarily diverse character of lower plate rocks, of which very few resemble the Franciscan. We conclude that an original subduction system was modified by later tectonic activity so that a variety of terranes was juxtaposed as a family of rootless thrusts, with the ophiolite forming, at least in some areas, the uppermost structural unit. Perhaps the emplacement of Wrangellia, an allochthonous microcontinent west of the San Juan Islands, caused the thrusting.

The Chiwaukum graben in central Washington contains two unconformity-bounded fluvial and lacustrine units, here named the “Chumstick” and “Wenatchee” Formations. The Chumstick is dated at 45 m.y. (middle Eocene), and the Wenatchee is dated at 34 m.y. (early Oligocene) by the fission-track method on zircons from tuffs. Previously, both formations were thought to be part of the Swauk Formation, which is older. The Chumstick Formation rests on weathered crystalline basement and is several thousands of metres thick. Fanglomerate occurs at the base and along the margins. Most of the formation consists of feldspathic sandstone and pebbly sandstone of fluvial origin; but within the upper part, there is a lacustrine unit, herein designated the Nahahum Canyon Member. Tuff is common in the lower part of the formation. Both of the bounding faults, the Leavenworth and the Entiat, were active during deposition of the Chumstick Formation, but relief was greatest on the northeast (Entiat) side. The Wenatchee Formation occurs in the vicinity of Wenatchee, Washington, and is ≤300 m thick. It unconformably overlies the Chumstick Formation within the graben but overlaps the northeast side of the graben, where it lies directly on weathered metamorphic basement. The Wenatchee Formation, like the Chumstick, was deposited primarily in fluvial and lacustrine environments; unlike the Chumstick, much of the sediment is mature quartz sandstone. Relief in the source area probably was very subdued during its deposition.

Aragonite and prehnite are widely associated in low-grade sedimentary, volcanic, and metaplutonic rocks of the San Juan Islands, Washington. The abundance of prehnite and the apparent absence of the high-pressure mineral lawsonite were thought to indicate that the aragonite had formed metastably. Our work now shows that lawsonite, commonly turbid and with a fibrous habit, also occurs in some of these rocks. The discovery of lawsonite indicates that the San Juan aragonite is probably a stable phase that indicates high-pressure metamorphism. The association of aragonite and prehnite also implies that the stability range of prehnite may extend to pressures above those of the calcite-aragonite transition. The occurrence of lawsonite is compatible with the interpretation that the bed rock of some of the San Juan Islands was subjected to deep and rapid tectonic burial by subduction at a convergent plate boundary.

Glacial and probably fluvioglacial rocks are present in the Currabubula Formation, an 8500-foot unit of Carboniferous age (Westphalian) exposed in the Werrie Basin of the New England geosyncline, New South Wales, Australia. Approximately 70 per cent of the formation is conglomerate and coarse conglomeratic sandstone, approximately 10 per cent is mudstone and fine sandstone, and approximately 20 per cent is tuff. The Rosedale Member, about 60 feet thick at the type locality, occurs in the middle of the Currabubula Formation and consists largely of varved mudstone with abundant ice-rafted pebbles. The best evidence for Carboniferous glaciation in eastern Australia is from the Rosedale Member and similar units exposed in other areas of northeastern New South Wales. The most common rock types in the Currabubula Formation are poorly sorted pebble conglomerates and lithic wackes composed largely of mineral grains and lithic fragments from flows and tuffs. Nearly all such rocks are stratified to some degree, and were probably deposited by streams, perhaps as glacial outwash. Rhyolitic and andesitic tuffs were deposited subaerially, in shallow lakes, and as ignimbrites. The Currabubula Formation was probably deposited near a mountain range undergoing glaciation. Much of the clastic material is volcanic, and includes pyroclastic fragments similar in composition to the contemporaneously deposited tuffs. This suggests that the clastic sediments were locally derived, and supports the idea that glaciation was restricted to the alpine variety. It is not necessary to infer either widespread Carboniferous climatic cooling or a relationship with the more widely occurring Permian glacial rocks in other parts of Australia.