Your browser is not supported. Please update it.

8 June 2018

3DFS talks of ‘software-defined electricity’ as DoE claims 62% grid wastage

R&D spin-out 3DFS has kicked the hornet’s nest, upsetting many in the energy community with something it is calling ‘software-defined electricity’ – or rather, a way of optimizing delivery to prevent wastage, thereby improving distribution efficiency. It comes as the US DoE revealed an estimate that around 62% of US energy is wasted through converting it to electricity.

Electricity is most efficiently transported at high voltages, and so it needs to be ramped up to these levels and then consequently transformed back down to a voltage suitable for use in a home – changing from AC in the generation into DC for use in the home. This process introduces some losses, around the 6% mark in the transport process, but it is essentially a cost of doing business in the current market – which uses large centralized power plants and the high voltage grid to distribute it to customers.

But the new DoE figures suggest that those ~6% losses in the transmission process are comparatively miniscule, as it estimates that 23.24 quadrillion BTUs (quads) are wasted as ‘conversion losses,’ in the production of electricity itself. This means that of the 37.7 quads of energy (from coal, gas, nuclear, renewables), 62% of this energy is wasted before there is even usable electricity. The DoE then estimates that of the 37.7 quads, around 2.2% (0.84 quads) is wasted in transmission and distribution (T&D).

3DFS contests that the conversion losses the DoE estimates are not as central as they might seem, rather that they are spread across the grid. The politics of this mean that the distribution community is pretty sensitive about allegations that they are collectively wasting electricity once it is in the grid, as these are losses that could be avoided. So, when 3DFS comes in with a revolutionary-sounding technology that could cut these losses, their hackles are raised.

Using edge-computing devices to monitor the flow of electricity, and a host of analytics software and algorithms to determine the best course of action, the 3DFS devices will inject tiny amounts of power into the flow – to optimally smooth out delivery to (nearly) perfectly match demand. This approach captures a huge amount of data, and is a huge opportunity for chip makers, networking specialists, and of course, energy suppliers.

The phrase ‘conversion losses’ somewhat implies an inevitability – that the process of burning something to create steam to drive a turbine to generate electricity has an unavoidable inefficiency. While true, 3DFS is arguing that there is a lot of electricity to be saved elsewhere in the supply chain.

The key to the 3DFS approach is a lot of edge-computing, to provide as near to real-time views of electricity as possible, so that the supply can be manipulated – matching voltages and demand with currents and supply. The startup argues that at scale, its system could double the efficiency of electricity generation – and that kind of claim is what has apparently upset the energy community so.

Terms such as “operating system for electricity” aren’t exactly helping counter the skepticism, but the 3DFS approach can definitely be called a management system – looking to exactly match supply with demand so that surplus electricity doesn’t get wasted as heat or inefficiently re-routed, or that an unexpected change in supply or demand doesn’t cause an outage.

As mentioned above, there is a lot of hostility towards the 3DFS claims. In this Vox profile, the writer recalls an irate professor of electrical engineering hanging up. That Vox piece is well worth the read, but profiles 3DFS as a company that is ahead of the current conventional understanding thanks to the data it is collecting and processing – “such that basically nobody beyond 3DFS has experience with them.” It also adds that the company has been around some 15-years, and has attracted a lot of talent.

The SDE term is probably more aggravating for the electrical engineers out there, but the goal is to smooth the supply to exactly match demand – so that devices aren’t having to adjust their own performance based on the good or bad quality of energy the receive. The thesis is that at the end-device level, these tiny fluctuations add up to a very big number of total wasted electricity.

Digital measurement is apparently key to this new way of doing things, instead of the analog-based Root Mean Squared (RMS) estimation used by most meters. 3DFS promises to be able to measure every cycle of the electricity’s sine wave, as it is delivered, so that it can see exactly (not approximately) how the juice is used. It claims to be able to measure 26 separate parameters in real time – adding up to over a million data points each second.

This data feeds 3DFS’ algorithms, although apparently about 99% of this data is discarded. These calculations are being run on what 3DFS calls Task-Oriented Optimal Computing (TOOC), and then those calculations form the basis of the corrections that are carried out by the Flash Energy Storage System (FESS), which adds or removes micro-amps of electricity to smooth out those sine waves.

The plan is to have 3DFS devices installed at end-points in the grid and at the customer premises. These boxes are called VectorQ, and have been deployed in data centers – one of the first areas that 3DFS is targeting, with Freudenberg IT a named customer. The company says that the system can achieve 98% efficiency, whereas a conventional set up would only be achieving 20-40%.

The VectorQ boxes are not cheap. Costing around $100,00, and the size of a wall-mounted networking cabinet, Freudenberg IT estimates they have an ROI of 7-years. However, the Freudenberg deployment showed cooler servers and lower noise in its data centers, which will compound the pure savings from energy expenditure. The price of the VectorQ family is expected to fall, as manufacturing scale increases.

Maritime ships are another area that shows potential, with micro-grids and distributed renewables also showing promise. The company is looking to miniaturize the technology so that it could be integrated into end-devices themselves, and it believes that it could be transformative for batteries too – helping to balance the load to reduce degradation.

The company is wary about the technology’s route to market, concerned by any investor or partner that seems too interested in its IP. The current plan is to focus on ‘behind-the-meter’ applications, those that the end-customers can control, before convincing utilities to invest in the grid-side applications. 3DFS is looking at organic growth, and says it is sustainable. The technology it has developed appears truly revolutionary, but it has a very, very, big hill to climb in order to achieve the energy savings it thinks are possible.