A key to expanding older DC systems with new equipment

Expanding existing DC power plants has been a natural part of telecom operators life for a long, long time. In later years the need to expand has become more frequent, at the same time as performing the expansions has become more difficult. The increased need to expand is due to the increased pace of change in the telecom network. Economical and technical life for telecom systems, which used to be several decades, is now considerably shorter. Frequent addition of new services, with added equipment, and colocation of for instance radio base stations are examples of things driving changes and/or increased power needs at existing sites. The operators want to accommodate this without scrapping fully functional existing power systems. They in most cases want to expand the plants already in place. At the same time the increased price pressure on power systems is driving an increased pace of new product introductions. The restructuring of the power industry has also led to power vendors both merging and folding the business. When an operator needs to expand an existing power plant, the likelihood of that particular system still being in production is considerably less today than for only a few years ago. It might even be that the supplier is no longer around. This has brought an increased need to expand existing power systems with newer generations of equipment, maybe not even of the same make as the existing plant. The major challenge in such an expansion is load sharing between the rectifiers in the new expansion and the existing plant. Unequal sharing puts unnecessarily high stress on one of the two parts, resulting in shorter life and increased risk of failure. There is even a risk of oscillation, further increasing stress and the consequences thereof. The paper will concentrate on a method of overcoming these load sharing difficulties, paralleling a new power plant with an existing, allowing the two plants to perform as one. This can be used to expand plants of all types and makes. The method is based on the new plant measuring the output voltage from the old plant and use that measurement to govern its own stable regulation, leaving the old plant to handle instantaneous loads variations. The result is stable operation from idle to full load for the combined pl- ant. The paper is describing this method, as well as giving examples of how it has been applied, including field experiences.