How machine learning is informing human understanding of energy reduction potential.


November 15th, 2017

There’s a lot of noise and confusion out there, we know the story… we hear it all the time: it starts with questions like: “Generation… can I still get a return?” “Battery storage… is it really that expensive?” “Tariffs… whatever happened to them?” And we know that sometimes you may be tempted to hold back on big decisions or just get someone in who can sort it all out; you’ve probably heard the line: “We’ll make you some money, it’s easy… just let us control your assets and install a battery”.

While some of that may be true, if you’re really interested in finding out the facts behind renewables and storage – and how they provide a sound financial return, please read on and if you can, join us at EMEX 2017 as we’d love to show you how our machine learning and data analytics is helping us to inform our client’s human understanding of investing in energy solutions.

Generation, Energy Storage and Analysis

The two big questions on everyone’s mind at the moment are; is it still possible to get a return on renewables and is energy storage really viable?

There’s so much noise concerning these 2 questions that trying to make an investment decision is really difficult.

So, should you jump in and invest in solar & storage?

We’d love to say that the answer to both is a simple YES.

However, the objective answer for an investment decision for any industrial, commercial, community or residential electricity user is a complex combination of their real-time demand & generation profiles, tariff pricing, location & the capital cost of the different solutions.

At Argand Solutions we’ve developed the analytical tools to enable you to make these tough decisions simple – filtering the noise and providing you with the clear factual answers using machine learning & energy analytics. Here’s how we do it…

Utilising machine learning and data analytics we’re able to rapidly analyse the optimal investment solution(s) for a single, regional or global portfolio of sites. Essentially, leveraging the power of big data we are able to analyse scenarios in minutes that would otherwise have taken weeks if manually analysed using excel or an equivalent package. By using your actual, modelled or scaled profile demand data we then overlay this with predicted locational generation data for a multitude of solar photovoltaic (PV) options and then characterise charge / discharge strategies for every time period (typically 30 minutes) for each of the chosen battery storage options.

With this process we’re able to run huge numbers of permutations to determine the optimal outcome(s).

In addition we can add in debt / equity financing structures and location specific export limits that may apply to a site to super-charge the financial analysis. The models outcomes can be based on internal rates of return (IRR), paybacks, cost of carbon, energy reductions & / or other metrics that are important to you and your decision-makers.

We’ll have this model running on the stand at EMEX 2017.

To see how this translates let’s look at an example for a large commercial building in Exeter, with a peak load of 150 kW and an overnight baseload of 50kW.

Here are the facts:

We modelled over 20 different battery options and 6 different solar PV options (from 0 to 500kWp) giving over 100 possible scenarios. The client wanted to achieve at least a 10% IRR and maximise the energy reduction at the site. Here’s their optimal solution:

  • Cost: £276k
  • Equity(%): 100%
  • Payback term: 8 years
  • Reduction in energy demand: 2%
  • IRR: 1 %
  • Annual savings (yr 1): £26.4k
  • Emissions reduction: 111T CO2 / annum

We can provide further clarity and choice by adjusting the model to suit your objectives.

Join us at EMEX to look at the facts and to see how we can remove the noise and confusion from making significant reductions and a return on your investment – right now.

EMEX takes place from the 22-23 November 2017 at ExCeL, London.