Overhead systems are out in the open, so it is easy to detect and fix design and installation problems. Underground problems, however, are out of sight and out of mind, at least until cables start failing. Although utilities design their underground circuits for a 30-year life, improper installations often can lead to premature field failures.

Unless you lay your cables to rest properly, they may come back to haunt you. Here's a brief example. A wind-generating farm was installed with underground cables tied directly to a main feeder cable. Unfortunately, the cables were simply placed in a trench using native soil backfill with minimal soil compaction. Ampacity calculations were performed using typical soil values, but thermal properties were not measured. Since wind turbines operate almost continuously, the feeder cable often ran at maximum capacity. The heat generated from the feeder cable dried out the surrounding soil completely. Because the native soil was poorly compacted fine silt, it acted like an insulating blanket and the cable failed prematurely.

A significant source of potential problems with underground circuits is the improper selection and installation of thermal backfill materials. To prevent premature failures, you must ensure you place cable systems in a hospitable environment.

Too few utilities have stringent specifications or quality-assurance programs for installing cable-trench backfill; this often leaves the decision up to the civil contractor. The effects of poorly installed thermal backfills and soils may not be evident for many years, until cable loads increase and temperatures rise beyond allowable levels, resulting in cable failures. The remedial cost of removing and replacing poor backfills is high, especially under paved roads. The loss of revenues from derating a system may be even higher. Installing a new circuit may be the only, albeit expensive, option.

All the heat generated by an underground power cable must be dissipated through the soil. This is quantified by the soil thermal resistivity (or thermal rho, °C-cm/W), which can vary from 30 to 500°C-cm/W. Electrical engineers understand the performance of the cable quite well, but to most, the soil behavior is a mystery, usually handled by using a thermal backfill with a supposedly “safe” thermal rho.

The ability of the surrounding soil to transfer the heat determines whether an operating cable remains cool or overheats. Improving the external thermal environment and accurately defining the soil and backfill thermal rho commonly results in a 10% to 15% increase in cable ampacity, with 30% improvements noted in some cases.

You can address potential problems by measuring the native soil's thermal properties and by using properly designed and installed corrective thermal backfills in the cable trench. In recent years, we've learned that using thermal probes connected to a Thermal Property Analyzer (EPRI EL-2128) can accurately measure the thermal rho in the field and laboratory.

The use of a soil thermal rho of 90°C-cm/W has become ingrained in cable engineering practices. Soil studies performed in the 1950s found this was a “safe” value for most moist soils. This value is commonly used for distribution cables, where cable loads are usually low and the native soil is used as the backfill. For transmission cables, it is assumed that the “thermal backfill” placed around the cables will be much better than the native soil and that it will have a thermal rho of less than 90°C-cm/W.

Most moist soils (with the exception of organic clays and silts, volcanic soils, peat and fills with ash and slag) have a rho of less than 90°C-cm/W. Moist sands, which are commonly placed around transmission cables, may even have a rho of less than 50°C-cm/W.

The critical word is “moist.” Many soils, especially uniform sands, can dry substantially when subjected to heat from the cables. The thermal rho of a dry soil would exceed 150°C-cm/W, and possibly approach 300°C-cm/W for a dry uniform sand. (The dry thermal rho of a properly designed and installed thermal backfill should be less than 100°C-cm/W and possibly as low as 75°C-cm/W).

In fact, a contractor, if left to his or her own devices, most likely would use readily available fine sand or concrete sand as the backfill. From a construction viewpoint, this sand makes an inexpensive and excellent bedding material, but thermally, it is very poor because it dries out easily under high cable loads.

Unfortunately, over the years utilities have used many unsuitable sands or “thermal backfills” because of ease of installation and availability. Several route thermal surveys of existing circuits installed before 1980 confirm this practice. Almost any sand, when moist, will give a reasonably low thermal rho. The crucial aspect is how easily it dries when subjected to cable heat loads.

Soils in semi-arid climates are naturally quite dry, so the assumption of a moist soil is not valid. It doesn't take much to dry these soils completely. In many parts of the country, the soil mineral and consistency is such that there is a high intrinsic thermal rho. Soil that is not properly compacted in the cable trench will be less dense and have a substantially higher thermal rho. Even distribution or low-voltage cables that are continuously under full load may dry the soil.

Cables that are near other heat sources, such as steam mains, will experience higher ambient temperatures, and if in the vicinity of other cables, will experience mutual heating and run hotter.

The thermal rho is important not only for transmission cables but also in any situation resulting in high heat generation. The assumption of a soil and backfill thermal rho of 90°C-cm/W may be erroneous, possibly leading to long-term problems when the cable is heavily loaded.

Poorly compacted trench backfill is a major problem. Not only is the thermal rho of uncompacted soil significantly higher, but the loose soil will dry more easily, which increases the possibility of thermal runaway.

Generally, native soils do not make good thermal backfills because their thermal rho values are poor, or they are difficult to properly re-compact in a cable trench. There are also problems associated with stockpiling, screening of debris, and contamination of good soil with organic topsoil. In the long run, the operational reliability gained by placing a classified thermal backfill around the cable has advantages over the variability and uncertainty of recompacted native soil.

Compacted granular backfills can have good thermal properties. Since most of the heat conduction is through the soil mineral particles and their contacts, one must ensure a high-density soil mixture to maximize these contacts. A well-graded sand to fine gravel can be a good thermal backfill when compacted to its maximum density as determined by a laboratory standard Proctor test (ASTM D698). The total cost of a compacted backfill must include material and transportation costs, as well as installation labor and quality-assurance costs.

The one often-neglected factor about compacted backfills is the need for quality assurance during installation. If the gradation of the backfill is not correct (sieve analysis ASTM D422), or it is not at the optimum moisture content (ASTM D698), or not enough compaction effort is applied, or the backfill lifts are too thick, then the maximum density will not be achieved and the thermal capability degraded.

Cement stabilized sand frequently has been used as a cable trench backfill in many countries. A typical mix design consists of 14 parts sand to one part cement, mixed with about 8% water. If the correct sand is used and properly installed, this material can have acceptable thermal performance. However, this backfill is quite strong and thus would be difficult to excavate. Quality control is required during mixing and installation, otherwise the thermal performance cannot be assured.

Many North American utilities have been using stone dust or crushed stone screenings as thermal backfill. If well graded and of the right mineral type, it provides a low and stable thermal resistivity when compacted at optimum moisture content and density. It does require thorough testing to establish density, moisture and thermal performance, and a good quality-control program to ensure proper installation.

With compacted soils, maximum soil density is needed in the restricted trench areas near cables or around cable pipe groups where proper compaction is difficult. Yet, it is precisely in these zones adjacent to the cables, where the heat flux is highest, that suitable compaction is most important to ensure maximum heat dissipation from the cables.

Over the past 10 to 15 years, we've seen great acceptance of fluidized thermal backfills (FTB^™), which are formulated to meet thermal resistivity, thermal stability, strength and flow criteria. This free-flowing, controlled-density fill is ideal for hard-to-access areas, such as narrow trenches, small diameter tunnels or areas congested with many underground services — basically where mechanical compaction is not feasible or practical. While the material cost of FTB may be higher, it should be considered for general usage because of its assured quality and quick installation, thus speeding up construction and decreasing overall costs, which are important factors when working in busy city streets.

FTB is a slurry backfill consisting of medium aggregate, sand, a small amount of cement, water and a fluidizing agent. FTBs can be formulated using locally available aggregates. The component proportions are chosen by laboratory testing of trial mixes to minimize thermal resistivity and maximize flow without segregating the components.

Be wary of commonly available “controlled density fills,” “flowable fills” or “slurry backfills,” which use large volumes of fly ash or sand. These may meet the mechanical and flow requirements for trench backfilling, but too often they provide totally unsuitable thermal performance. Fluidized thermal backfills should be formulated and tested only by soil thermal specialists who understand the tricks of the trade in making thermal measurements.