Entropy is a fundamental concept that applies to all nature, including human activity. There are various ways to describe the concept of entropy:
The last statement is the one of most importance in trying to achieve sustainable development. In a development context the "system" should be defined in the usual way as the part designed and built by humans and include the environment to which it is unavoidably open. To have low-entropy (sustainable) development it is necessary to redefine the "system" to be designed and built to include as much of the environment as possible. The ultimate environment for humans on Earth is the Sun. If humans are not to harm the Earth's environment, they must design their society to use the entropy increase of the Sun, not the entropy increase of the Earth's environment.
How humans increase Earth's entropy:
Components of useable solar entropy are:
Keeping the entropy low for the defined system is not the only consideration in sustainable development. A very low entropy state is one of identical molecules in a matrix equally spaced apart; not a viable development situation. In addition to low entropy one also needs a certain degree of complexity or diversity. (See Signs of Life: How Complexity Pervades Biology.) One only has to look at nature to see low entropy and complexity work together. (See Biomimicry.) What is needed is low entropy and high complexity. (A complete mixing of many types of molecules would be high entropy and high complexity.) However, high complexity is more difficult to manage both initially and finally for different reasons (see below). The lesson from biology is that highly complex systems manage themselves; when one small subsystem fails, other subsystems take up the slack.
Complexity experiments and theory (Signs of Life: How Complexity Pervades Biology) show that there is a "sweet point" at which the degree of complexity of a system causes a "phase change" in the system from a system made of parts to a self-organizing system that is more than the sum of its parts. This is due to the non-linear nature of the interactions between the parts and feedbacks among parts. Biological systems benefit from such a sweet point. Systems beyond the sweet point are constantly changing to adjust to changes in the environment surrounding it.
A system of human development cannot completely mimic the low-entropy/high-complexity of biological systems, because it has the goal of providing goods and services for the benefit of humans, not the entire biosphere. A low-complexity development might provide short-term human benefits, but usually deleterious long-term effects for humans and the rest of the biosphere. Increasing the complexity by adding more closely linked activities to the system might reduce the amount of short-term human benefits, but generally will increase the long-term benefits. It certainly requires more research and design effort to plan such closely linked activities and surely will require more effort in administration, at least initially. Eventually high enough complexity should provide some of its own management, perhaps thereby decreasing the human management needed. At some point as the complexity is increased the "sweet point" may be reached at which the system becomes self organized, perhaps reducing the effort required to manage it. A human development system beyond the sweet point will be constantly changing to adjust to changes in the environment around it. These spontaneous changes may not always be directed toward providing goods and services for the benefit of humans, but instead for the system as a whole. As the system develops in complexity some management tweaking may be necessary to keep the goal to provide goods and services for the benefit of humans.
Perhaps it is possible that human societies will become complex enough in the future such that the societal goals will move more toward providing benefits for more than just humans. That would certainly be a major "phase change" due to the level of complexity, since human social evolution has mostly seemed to advance in the past by humanity working toward its own benefit.
Industry (including agriculture) uses low-entropy/high-complexity Earth minerals, including energy minerals, and low-entropy solar energy (crops, forests, solar heat, water, hydro-energy, wind energy and solar-radiation energy) to produce low-entropy/higher-complexity products and higher-entropy wastes, with the total entropy always increasing in the process. Some of the low-entropy/high-complexity minerals involved low-entropy solar energy in the far-distant past (mineral fuels, e.g., coal, oil and gas) and some of it involved solar energy and other radiant and mechanical energies of the galaxy (e.g. metal ores).
The low-entropy/high-complexity minerals are finite and small in extent at current usage rates when compared to the likely future time of humans on Earth. (See http://www.roperld.com/science/minerals/minerals.htm.) The low-entropy solar energy is much larger and always available in the same comparison. Thus, it is wise for humans to use low-entropy/high-complexity minerals sparingly and mostly as matter instead of converting it into energy, using mostly solar energy in its various forms to reshape and recycle the matter.
As low-entropy/high-complexity minerals are mined, what is left to mine at any time is higher in entropy than at previous times. For example, a specific metal ore requires more inputs of materials and energy to mine and refine it to the metal as time goes on. Eventually it costs more in materials and energy to mine and refine it than is achieved by mining and refining it. Of course, eventually all the mined minerals will be fairly uniformly scattering around Earth, and thus will be unavailable for further human use; then only energy provided by the Sun will be available for use by humans for primary energy and for creating useful materials by means of the biological cycle.
Put in terms of closed and open systems, using mainly mineral resources for both matter and energy in industrial processes is relying on a closed-Earth system. Using solar energy in its various forms for the energy, and mineral resources mostly for matter instead of energy, is opening up the Earth system to include the sun more than at present. In the latter case the part of Earth involved in the industrial processes does not increase in entropy as much as it does in the former case. (The entire Earth decreases in entropy as the sun increases in entropy, with the sum of the two increasing in entropy, when neglecting other entities in the solar system and galaxy.)
The ultimate unachievable goal for sustainable development is to waste nothing (matter/energy) in the defined system. The goal is unachievable because of the Second Law of Thermodynamics: some matter/energy will always escape into surrounding systems. So it comes down to reducing wastes in the defined system as much as possible.
Cluster industries, where some of the "waste" output of one industry is used as part of the input for another nearby industry, as well as sharing labor and infrastructure, is a good way to keep the increase of total entropy of the cluster to a low value. (Do not confuse this with a common definition of "cluster industries" which refers to industries of the same type being clustered, such as electronics industries.) In general, waste-reducing cluster industries will involve quite different types of industries.
The best type of cluster would be where the industries form a complete cycle or, more likely, several complete cycles, in which the wastes of any industry in the cluster is used as a main item of input by some other industry in the cluster. (This is the way nature works; so such clustered industries would be practicing biomimicry.) The industries in a cluster would not have to be physically located next to each other, but ideally should be in order to cut down on transportation entropy increases. Of course, even more ideal would be for the products of the industries to be used locally. The ideal case is a self-sufficient community except for solar input.
Ideals are never realized. So, are there examples of cluster industries in the real world? Here are some (called Eco-Industrial Parks) in the United States: http://www.p2pays.org/ref/06/05505.htm#iv. See Gunter Pauli's book Upsizing and his on-line article, No Waste Economy, for many examples around the world.
Gunter Pauli in his book Upsizing has extensive discussions about opening up industrial systems by including as many of the five kingdoms of life as possible. A specific example is Brewing a Future.