Upgrading Belmont’s Grid

One of the important questions we face as we navigate towards an electrified future is the need to upgrade local power distribution systems.  The need to upgrade regional transmission systems has gotten a lot of attention.   The costs of upgrading local substations, wiring, and transformers have gotten less attention. 

Recently, I sat down with Craig Spinale, the General Manager of Belmont Light — the municipal utility that distributes power in Belmont.  He gave me a very rough sense of the kinds of investment that would be needed to support full electrification of heating in the town.  He was kind enough to allow me to publish these notes from our conversation

Belmont’s Power Distribution System 101

Belmont Light serves 10,949 residential meters and some commercial, industrial, and municipal meters.

Belmont Light receives power through a 115KV line coming in from Cambridge to the substation at Blair Pond.   At that substation, two 60MVA transformers step the voltage down to 13KV.  The highest summer peak load experienced has been approximately 36MVA.  While there are two transformers that could theoretically handle a load of 2 x 60MVA or 120 MVA together, that would not be an ideal condition because if one went down, a significant part of the town could be out for an extended period — the idea is to keep one transformer in reserve.

From that substation, ten 13KV lines emanate across the town.  They can carry 8MVA each.   Five of the lines go out across the town and directly serve local transformers.  Another five of those lines feed older substations that step down the voltage further to 5KV and deliver power to areas of the town are still on the 5KV service.  It is a work in progress to eliminate the old substations and directly serve the whole town with 13KV lines.

Homes in the town need 240/120 volt service.  So, there are additional transformers that tap into the main lines (at either 13KV or 5KV) and step the voltage down to 240/120 service.  There are approximately 900 of these transformers serving the 10,949 residential units in the town.  The current standard size transformer is 50KVA and with traditional load expectations of up to 3KVA per home, they can handle 10 or 20 homes.  However, many of the transformers in town are older and smaller at 25KVA and can handle only 5 or 10 homes.  The traditional assumption was that homes would not simultaneously draw more than an average of 3KVA.  With electrification, expected load my run much higher and with battery and solar power discharge available, the load may go in either direction.

Roughly 2/3 of the feeder lines (13KV or 5KV) distributing power around the town are in underground conduits. 

Upgrades required for full electrification.

  1. Approximately $5,000 or 10,000 per home for transformer upgrades. At the level of the individual home, with a combination of heat pumps, batteries, solar and electric vehicles, many homes are choosing to upgrade to 200 or 400Amp services because their home could draw several dozen KVA  or could send several dozen KVA back to the grid.  Especially for a 400Amp circuit, the home may need a dedicated transformer; the home will certainly need to be on a transformer serving fewer homes.  The cost of installing a dedicated transformer may be up to $10,000 or even more (excluding the roughly $5,000 panel upgrade that might be required inside the home).  Depending on where the existing transformer is located, it may be necessary to reconfigure the connections from the 5 or 13KV feeder lines along the block.  A new transformer could also force the replacement of the pole that would carry it.  Replacements of transformers in areas where power is underground may involve additional costs to create  appropriate manholes to install the switches between the transformer and the feeder lines.  Currently, BMLD usually bears some of the cost of the transformer upgrade but charges the customer for much of it.
  2. Approximately $100,000,000 for an additional substation.  It is expected that in a fully electrified home, the winter heating load will peak at roughly 3x the current max load for air conditioning.  During a cold snap, all heat pumps will need to be running at simultaneously for an extended period with limited potential for demand management after the first day or so of the snap.  That would suggest a peak load approaching 100 MVA.  The current substation cannot handle more than 60MVA with full reliability.  Additionally, the ten lines coming out of the substation can only handle approximately 80MVA (10x8MVA).  To meet the new peaks and to provide appropriate reliability and redundancy it will likely be necessary to bring in a new 115KV transmission line to a substation at the west end of town.  The recent new east-end substation cost approximately $100,000,000. The need for an additional substation is a decade or two away.
  3. Other potential costs.  The 13KVA feeder lines running around the town may need upgrades where they are underground.  Lines on poles can dissipate heat easily, but underground lines need to be sized to stay cooler.

With 10,949 homes in town, this cost of upgrade could work out to $20,000 per home — roughly $1000 per year per household for the next couple of decades if the costs were socialized across the rate base.  All of these numbers are very approximate, and could be too low or too high.

Return to heat pump outline

Published by Will Brownsberger

Will Brownsberger is State Senator from the Second Suffolk and Middlesex District.

10 replies on “Upgrading Belmont’s Grid”

  1. “ roughly $1000 per year per household for the next couple of decades if the costs were socialized across the rate base. ”

    Some questions to consider. Perhaps there are more.

    What additional amount would Belmont Light have charged annually for the ‘normal’ upgrades required without the investment for the substantial all-electric upgraded?

    What would residents save annually for the asserted lower cost of all electric heat pump and water heating?

    An added cost not stated here: what would the “financing” cost be annually for each household, a financing cost incurred by Belmont Light?

    What would be the household cost of ‘doing nothing’ (for a better climate)? For example, more air conditioning units and more electricity will be a cost of doing nothing.

  2. Obviously with a potential capital expenditure that large, Belmont Light needs to start now to map out the more intra-system issues that will drive the need for extra capacity and whether there are alternative ways to address the reliability and risk concerns. They mention peaks as a big issue; are they system-wide or more localized? And what role can prudent investments in storage play in solving that problem at much lower cost? What portion of the reliability concern can be addressed through a combination of storage, more local generation, and load shifting to off-peak hours?
    The analytical work has hopefully already begun within BMLD; within a year or two they ought to be presenting any early investments they believe will help them to help delay or avoid the need for this investment.

  3. Thank you for this example of Belmont, Will
    1. Presumably then there will be a decreasing and eventually a zero need for gas. This example illustrates the imperative of integrated energy planning and not letting Keyspan (National Grid) proceed with its horrifically expensive and for now effectively unmonitored GSEP program and intention of delivering gas into the indefinite future. Instead National Grid and other gas utilities should be initiating a program of strategic retirement of their gas systems, with in parallel authorization (as proposed in SD.2330) to supply thermal energy not gas alone and to participate in heat pump-related business opportunities.

    2. This analysis also points to the need for and value of demand management techniques, reflected in the restructuring of electric rates so demand peaks are flattened, since unlike heating many demands for electricity can be deferred or scheduled away from peak periods, e.g. EV charging, use of appliances such as washing machines, dryers, encouraged by time-of-day and load-sensitive rate structures. The role of local or neighborhood energy storage should be explored, if there are suitable locations and spaces for their installation.

  4. Hi Will, thank you for sharing the detailed review. I have appreciated your posts on energy matters. However, I am concerned that this post on Belmont’s grid could be misleading in some ways. for example:
    1) Currently, there is no lack of system wide capacity in Belmont for electrification, and my understanding is that we don’t have to worry about running out of capacity for at least five years. It would be unfortunate if a “long term” analysis dissuaded Belmont customers looking to electrify in the “short term” from proceeding.
    2) The cost of upgrading the grid to support full electrification will be substantial and I think your ballpark figures are great for initial planning. However, what I believe was left out was the also substantial growth in revenue for Belmont Light. With the additional sales of electricity (at the current electric rates), I could see Belmont Light tripling its annual revenue from around $25 million to $75 million, or possibly more. This additional revenue, and its ability to help pay for upgrades, should be taken into account when thinking about the financial impacts of an upgrade. There wouldn’t necessarily need to be an increase in rates to pay for the upgrades.
    3) A winter peak of 100 MVA seems high to me, and I have been skeptical of other values suggested for full electrification. Do you have a link to the analysis you were using for this? I have a hard time believing that value, especially as (1) EV’s can charge off peak, (2) thermal storage technologies are advancing to shift load off peak, and (3) district geothermal (and it’s much higher COP under extreme cold) may serve a substantial number homes under an all-electric scenario. In addition, I suspect there may be some double-counting in the peak load when it comes to distributing the heat in a home.

    Thanks for highlighting these issues, allowing me to make comments, and keeping the discussion going.
    Dave Beavers, Belmont MLB Vice-Chair

    1. Thank you for these thoughts, Dave, and thank you for your service on the light board.

      First, absolutely agreed that Belmont has gobs of room to grow at this time. The need to upgrade Belmont’s substation is two or three decades away. I certainly don’t mean to suggest that there is present cause for alarm. On the other hand, the grid improvements needed to support our desired end state of full electrification will take time.

      Second, also absolutely agreed that this is not a financial threat for Belmont Light. I’m taking the perspective of consumers and what we will together need to absorb in our bills one way or another.

      Finally, somewhat agreed as to the haziness of ultimate peak load projections, but I don’t think a winter cold snap peak three times greater than today’s summer heat wave peak is unduly pessimistic.

      The following considerations lead me to believe a 3x increase in peak load (shifting from summer to winter) is a scenario worth considering:

      Our stated goal is near total electrification of residential heating in 2050, so that is the premise of the analysis. Most homes will have air source heat pumps, since individual ground source heat pumps are not viable on the vast majority of Belmont lots. Agreed that there is some possibility that networked ground source pumps will be available to increase efficiency and that would be highly desirable, but that is still a speculative idea at this stage; the engineering and economics are very unclear.

      So the winter peak scenario we are considering is that most homes have air source heat pumps. By 2050, the global average weather will be warmer, but my additional premise is that a polar weather episode that keeps temperatures near zero for a few days will remain something we should be prepared to handle.

      In that case, a sustained peak three times greater than our current heat wave summer peak is a substantial likelihood. Consider the following as to heating load alone:

      • A heat pump’s power consumption is a function of how much heat it has to move, which is a function of how different the temperature outside is from the inside. So, heat pumps in A/C mode would have to lift across 30 degrees if it’s 100 outside in the summer. But they have to lift across 70 degrees if its 0 in the winter. So, the peak heating load is roughly double the peak A/C load for that reason.
      • But then factor in that heat pumps get less efficient at lower temperatures. For example, even with a good cold weather pump, realized efficiency could drop from 300% to 200%. So, for an individual home, the winter peak for heating would likely be three times the summer peak for cooling with central air.
      • And then consider that AC penetration today is only partial. In my recent survey on heating issues, I had 695 Belmont respondents. Among those Belmont respondents, 47% had partial AC and 13% had no AC — only 40% had central AC. So, today’s summer peak HVAC load is much smaller than it would be if everyone were fully cooling their homes. In the winter, everyone will need to fully heat their homes — one has to keep the pipes warm, not just the bedroom. So, in the winter cold snap the heating peak HVAC load might be 50% or 100% higher than the current summer HVAC peak due to penetration alone.
      • Put the load swing together with the efficiency decline and the penetration increase and you get a winter HVAC peak that could be 4x to 6x the summer HVAC peak. Since HVAC load is about half of the current summer peak load (see the residential baseline study), a 5x increase HVAC load would mean a tripling of total load.

      So, a winter peak 3x today’s summer peak is a reasonable ballpark, even before one starts to consider electrification of other home appliances and the introduction of electric vehicles. I agree that EVs will not necessarily contribute to peak, because they can be managed. But neither do I consider that behind-the-meter batteries or solar will carry us through a winter cold snap peak. Solar is always weak in winter and one cannot count on sunshine. Batteries at a household scale are very expensive and will not cover more than a day or two of heating load.

      In my mind, there are four things that could obviate the possibility of a 3x increase in peak load: (a) networked geothermal could prove out and get built, but that’s still very speculative; (b) homeowners could invest in deep energy retrofits, but these are much more costly per home than the grid upgrades we are talking about and are not likely to be universal; (c) the weather could just get much warmer and polar vortexes could recede as a threat; (d) due to cost and labor force issues we could fail to hit our goals for widespread heat pump adoption. I’m not quite prepared to bank on the latter two possibilities.

      1. Will, that was great response to my feedback. I feel privileged to have someone representing me in the senate that has the capacity to discuss issues like this in such detail. Thank you for all the good work you do.

        First, per the “gobs” of room to grow out electrification under our current distribution system, this appears to be a little-known fact around Town. I get asked about that frequently, and I sense skepticism when I say we have plenty of capacity. I hope this conversation will help get the word out.
        I believe your analysis is thorough and well thought out, and I don’t think it would be productive for me to quibble with any assumptions or calculations you did. Rather, to forward the conversation I’m going to take what you’ve produced as a “vision” of the future, and:
        1. Provide a friendly critique of that vision and explain why it doesn’t necessarily have an optimal outcome for Belmont.
        2. Provide an alternative vision – admittedly a bit absurd, but one that hopefully will make the point that we could have a better outcome with less transmission and substation capacity.
        3. Consider a path forward for planning to, as you say, navigate toward an electrified future

        (1) Friendly critique of a mostly air-source heat pump, 3X summer load for peak winter, scenario.
        a. Inefficient allocation of resources. This configuration would likely be inefficient in an economic sense. Today (2023) Belmont Light scores poorly on a metric called Load Factor – I recall a rate consultant saying that he had never seen one so low. This is the ratio of average load (kW) to the peak load (kW). The average load correlates with revenue that Belmont Light can generate, while the peak load is associated with the capital cost of the distribution system we need to serve peak load. In a nutshell, relative to other utilities, we pay for more grid capacity and generate less revenue. This is due to a higher percentage of our load being residential, and it pushes the electric rate we need to charge higher.

        The mostly air-source HP scenario would likely exacerbate the inefficiency in the current Belmont grid. While the peak goes up 3X, the shoulder months (e.g. May) when neither heating nor air-conditioning is happening, would not likely change that much. Thus I would not expect the average load to grow in the way the peak does, and this would make for a poorer Load Factor. Imagine the new, expensive, transmission and substation equipment sitting idle for most of the time – that’s not a good design from an economic perspective.
        b. Substantially higher transmission costs. The winter peak of 3X of summer peak would likely increase the cost of transmission services in the winter by a factor of 5 or so (i.e. the future winter peak will be something like 5X the current winter peak). This additional cost will be passed on to rate payers.
        c. Higher summer costs. The air-source heat pumps will likely be used in the summer for air-conditioning, increasing the summer peak and increasing summer transmission and ISO-NE capacity costs.
        d. Doesn’t help with resiliency. One grid planning element that is being discussed these days is resiliency. This is in the context of global warming and possible risks, but also attacks on substations or hacking of the New England grid are not out of the question. While moving towards reliance on one source of energy (electricity) this vision doesn’t appear to improve the energy resiliency of the Belmont grid.
        e. Spends a lot of local money (i.e. rate payer money) out of Town. At least the transmission part of the upgrade would likely entail paying Eversource a lot of money. This expenditure would not likely help local businesses or residents in an economic sense.

        (2) Alternative Vision – 100% Electrified Future with current 60MW BL system capacity
        OK, I admit this may seem absurd, but for reasons provided below, I think it should be considered at least hypothetically for planning purposes.
        Assumption: with 100% electrification our average load (across the year) goes up 3X to roughly 40MW.
        a. Obviously, there is very little room for a “peak” of 60 MW above the average load of 40 MW. Under this scenario extreme load-flattening would need to be employed.
        • Air-source heat pumps would need to be coupled with thermal storage and charge the thermal storage off-peak
        • A majority of homes would need a heat-pump technology that lifts across a 20 degree difference, not the 70 degrees you noted for an air-source heat pump on a cold day. Geothermal thermal is an obvious candidate. I agree individual home geothermal is not likely the answer, but I do see a future for district-geothermal. The technology is ready for the most part, it’s the business and regulatory side that needs to be worked out. With the pressure on National Grid to reduce carbon, and their interest in exploring district-geothermal as a business opportunity, I could see this happening.
        • EV’s obviously must charge off-peak.
        • Establish electric rates that pass on the true-hourly cost of providing electricity, so customers will have an economic incentive to avoid peak hourly rates
        • Battery – electric storage, both at the Belmont Light system level and home level could help in a complimentary way – but I agree not in a primary way
        • Cost effective energy efficiency – can supplement, but most effort would be directed toward load shifting technologies
        b. Benefits
        • Economically efficient. A much improved load factor with a better average load to peak load would mean a much more efficient use of distribution and transmission equipment
        • Savings (40% or more over the 3X scenario) on winter transmission costs for Belmont Light
        • Summer transmission and ISO-NE capacity cost savings for Belmont Light
        • Thermal and electrical storage will help improve the energy resiliency of Belmont
        • More of the $100 million or so investment could be spent locally improving homes and businesses
        c. Costs?
        • I don’t have a mathematical workup like you provided, and unfortunately my scenario would be more complicated to assess. I think the $100 million saved from avoiding distribution and transmission upgrades would be a start. But likely more would be needed. I expect there will be other resources available (e.g. National-Grid gas financing and operation of district-geothermal) and the benefits as-a-whole may well outweigh the costs.

        (3) Planning
        I apologize for taking your straight-forward analysis and complicating it. But I truly believe we should consider more than just the raw capacity needed for electrification, (and associated cost) when planning such a large and long-time project. As I hope I demonstrated above, the 3X scenario may not lead to the best outcome for Belmont. A planning exercise might include:
        a. Consulting with Belmont residents, Belmont Light, and others about the goals we may wish to achieve (e.g. in addition to 100% electric, an efficient allocation of capital – higher Load factor, spend as much money as possible in Town, spend the least amount of money, etc..)
        b. Develop scenarios (e.g. your 3X scenario could be one, and the no additional capacity scenario another, and some in between those extremes) and assess if they may be achievable and how they meet the goals identified
        c. Identify possible pathways and steps that would need to be taken for success in each scenario over the next 30 years

        I think such a planning exercise would likely require hiring a consulting firm. But there may be grants available (DOE?) for such planning exercises.

        I hope this has not been too long-winded. In a nutshell, in the same way that those looking to install solar PV are advised to invest in energy efficiency first, so that the energy generated by the high-end technology is not wasted, I believe we should look to optimize our grid operations with the current 60 MW capacity before seeking to upgrade the capacity. In doing so, I sincerely believe we will not need the level of upgrade you calculated to get 100% electrification, and we could receive more benefits from the effort.
        Thanks for addressing this topic and allowing me to offer comments.

  5. There is a complexity suggested in the comments above.

    The programs and analyses that may apply include:
    Availability of incentives for customers to electrify, comparable to those available to Eversource customers;
    Training of workers for the increased utility level of electric work, and the retail/heatpump work;
    Moderate income home ownership and rental housing support for these costs;
    To name some.

  6. There are 2.6 M residences in MA. If it costs $20k/residence to upgrade LOCAL, intra-town distribution wiring, then that requires $50B, most of which needs to be done by NGRID and Eversource. They may be more efficient by not being restricted by town boundaries, but we’re still talking major $$. And this doesn’t include increasing the ISO-NE Grid by ~3X – how much does that cost? And who will pay (ratepayers, of course)? Are NGRID, Eversource, and ISO-NE currently (sorry for pun) planning logistically and capital-wise for these kinds of investments??

  7. There is a system to evaluate the costs we are talking about, those described above by Mark. The system is called Cost of Service. It is employed by the MASS DPU. With public, scientific, and financial experience and input, the formal system to evaluate the costs and savings of ‘conversion to all electric’ could be enhanced to evaluate the current question. Once cost of service included the price of generation. Generation prices are now set competitively. Now it covers ‘delivery’ costs. The costs include, but are not limited to: transmission lines, large ‘step-down’ substations for example those bringing power from Hydro Quebec, distribution substations that convert power to voltages in the local overhead and underground lines, transformers on poles or underground converting to house and large building voltages, the costs of managing and repairing, and more like, the cost of metering.

    Shouldn’t the DPU open such a case for conversion to all electric?

Comments are closed.