Solar for wineries and vineyards: the off-grid and microgrid guide

Every figure on this page comes from a published source: a university or government study, a trade-press case study, a national broadcaster report or a manufacturer datasheet, each linked inline. Projects in the United States and France are labelled as such, they are evidence for what works, not RENOZ projects. Australian exemplars are labelled with their state. Nothing here is a substitute for a site-specific design by an accredited installer, which is who owns sizing, protection and commissioning on every project.

We also hold ourselves to the same verify-live discipline we recommend: no standing approval claims for any exact battery model, RENOZ included. Scheme eligibility and product listing status change, so confirm the exact model numbers on any quote against the current CEC approved-products list(opens in a new tab) before signing.

Winery electricity splits into four blocks. First, refrigeration: fermentation chilling during vintage, then barrel-room and tank temperature control all year, which is why refrigeration dominates the load profile in the UC Davis analysis of winery microgrids (2023). Second, the crush spike: presses, destemmers, pumps and cleaning run hard for six to eight weeks while fermentation chilling adds a round-the-clock demand block on top.

Third, hospitality: cellar door, restaurants and weddings, which is exactly how the Lowe Wines system in Mudgee is loaded, a winery plus cellar door plus restaurant plus 160-guest weddings on a fully off-grid site(opens in a new tab) (ABC, 2025). Fourth, irrigation: bore and surface pumps, and here the off-grid physics bites, a direct-on-line motor start draws five to seven times nameplate current for seconds, so the inverter's overload curve matters more than its continuous rating. That curve is the core argument of our off-grid battery guide.

A practical example of how the blocks interact: a winery that hosts weddings runs its biggest event loads on summer weekends, which is also when a weak rural feeder sags and when bore irrigation competes for the same supply. The UC Davis analysis of winery energy use notes refrigeration as the dominant load (2023), but the operational risk concentrates where those blocks overlap, and that overlap window is what the storage design has to carry.

**Crush** is the vintage harvest period when fruit arrives, is pressed, and fermentation begins, and it is the single worst load event a winery connection faces each year. Any solar design that averages the year and ignores the crush will fail exactly when it matters.

Here is the framing that most crush-led thinking misses: the crush is a power problem, the other forty-six weeks are an energy problem, and the energy problem is where the annual bill actually accumulates. Barrel rooms and tank farms hold temperature 365 days a year. Case-goods warehouses, offices, tasting rooms that trade every weekend, yard and cellar-door lighting, bore pumps, wastewater treatment, compressed air for bottling lines: none of it is glamorous, all of it is always on, and together it usually outweighs the six frantic weeks of crush in annual kilowatt-hours. That baseload is also the easiest money on the site: it is daytime-weighted, predictable, and a plain grid-tie array offsets it without any storage drama. The honest sequencing is therefore baseload first, then storage for the crush and the outage risk, not the other way round; get the year-round offset wrong and you have bought a resilience toy that never pays for itself.

**A winery microgrid** is a site-level power system where solar, battery storage, controls and optionally a generator can operate together and separate from the grid when needed. It is not a marketing term here, it is a specific architecture: the battery and inverter must be able to form their own grid, which is what separates a microgrid from ordinary grid-tie solar.

Grid-tie solar remains the right answer where the connection is healthy and the load is daytime-weighted; the US evidence bears this out, with the Korbel winery in California offsetting 90% of its average bill with 852 kW of PV and 360 kW / 798 kWh of storage(opens in a new tab) driven mainly by demand charges up to $22/kW (EnergyToolBase case study, 2022). But every US case study of this class is grid-tied, which is the gap Australian operators with weak rural connections face: if the wire is the problem, grid-tie does not solve it.

Full off-grid removes the wire entirely and inherits the whole duty. Hybrid microgrid sits between and is where most commercial agri sites land: keep the grid where it is usable, let the battery carry crush peaks and outages. The battery-versus-inverter architecture underneath this choice, 48V rack modules on a grid-forming inverter-charger versus closed high-voltage grid-hybrid sets, is covered in our 48V versus high-voltage guide; the same open-versus-closed logic applies at winery scale.

A caution from the field: closed grid-hybrid battery sets sold for residential self-consumption are usually grid-referenced, with backup as an event rather than permanent duty, and their published backup overload runs roughly 1.2 to 1.4 times rated output for 5 to 10 seconds (manufacturer backup datasheets). For a site with direct-on-line pump and compressor starts, that gap is exactly why the 48V open architecture argument in our off-grid battery guide carries over to winery scale. Ask any quoting installer to put the overload curve and the generator plan in writing.

For comparison shopping across the 48V-class field, PowerPlus publishes its LiFe4851 rack module details(opens in a new tab) and BYD publishes the Battery-Box Premium LVS range(opens in a new tab); both are peer vendors to RENOZ in the open 48V architecture, and the same verify-exact-model discipline applies to all three.

**A demand charge** is a network charge based on the highest power draw in a billing period, not the energy used, and it is why a six-week crush can distort twelve months of bills. Korbel's economics were driven by demand charges up to $22/kW (EnergyToolBase case study, 2022), and Australian network tariffs for commercial and industrial customers use the same mechanism under different names.

The sizing method is boring and non-negotiable. Pull 15-minute interval data for at least a full year so the last vintage is in the record. Separate running loads from starting loads, because motors that start direct-on-line demand five to seven times nameplate for seconds. Size PV to the worst month, not the annual average, and size battery autonomy for the outage or grid-limit scenario the business actually fears: one night of fermentation chilling is a different bank to three days of full crush.

One sizing trap deserves its own paragraph: interval data shows running load, not starting load. For instance, a 1.5 kW bore pump starting direct-on-line can demand 7 kVA or more for seconds (locked-rotor current, five to seven times nameplate), which is why the inverter-charger overload curve, not the battery kWh figure, decides whether pumps start on a cold morning. Consider sizing the worst-case simultaneous start before anything else, then work back to energy.

For bore pumping, US extension and vendor methods size PV at roughly 1.25 times the pump wattage and stack multipliers for inverter efficiency, temperature and depth of discharge, with two to three days of autonomy typical and up to five or more for remote sites (US extension sizing method). Our WA battery sizing guide works the same arithmetic step by step for WA conditions.

Lowe Wines, Mudgee NSW, is the Australian proof that full off-grid works at production scale: 265 kW of solar across 414 panels and 1.2 MWh of battery(opens in a new tab) carrying the winery, cellar door, restaurant and events with no grid connection (ABC, 2025). As stated arithmetic against RENOZ's published platform, 1.2 MWh is about five HC-125K-261-02B cabinets at 125 kW / 261.25 kWh each (RENOZ commercial platform; Lowe is a non-RENOZ project and that comparison is scale arithmetic, not a claim about their hardware.

Henschke in the Eden Valley, SA, took the hybrid path: 47 kW of existing solar plus 50 kW added, a battery and diesel backup(opens in a new tab) as a winery microgrid (Tandem Energy case study). Note the generator stayed, hybrid does not mean generator-free, it means the generator becomes an occasional insurance policy instead of the backbone.

The California evidence supplies the economics. Domaine Carneros in Napa runs 427 kW of solar, a 250 kVA battery and a generator(opens in a new tab) and supplied 70-80% of site load in 2024 (Microgrid Knowledge); their stated motivation was resilience, lose power at harvest and fruit overripens. A California Energy Commission demonstration of 200 kW PV with 280 kWh storage across two winery buildings achieved up to 39% peak reduction and about $12,000 per year in savings(opens in a new tab) (CEC-500-2022-006, 2022). Every one of these is a non-RENOZ, non-WA project; they prove the architecture, not the vendor.

Domaine Carneros also supplies a lesson the glossy case studies skip: their microgrid took about six months of controls fine-tuning before it behaved the way the design intended (Microgrid Knowledge). A winery microgrid is software and commissioning as much as hardware, so ask any quoting installer how the islanding, load-shedding and generator-start sequences get tuned and tested on your site, not just what the cabinets cost.

WA's wine map splits into two connection stories. The Swan Valley and Geographe sit close to the metro network, where the constraint is more often tariff structure and demand charges than raw capacity. Margaret River, the Great Southern (Mount Barker, Porongurup, Frankland River) and Pemberton sit at the end of rural feeders, where end-of-line supply, three-phase availability and capacity caps are the lived reality, and where the microgrid and off-grid options earn their keep.

The **South West Interconnected System (SWIS)** is the main electricity network covering Perth and WA's South West, and much of the state's wine country sits on its rural fringes or beyond it on stand-alone feeders. Away from the SWIS, stand-alone systems are designed under AS/NZS 4509 (Standards Australia(opens in a new tab), the same standards regime our flagship off-grid battery guide walks through.

We deliberately publish no specific SWIS capacity or kVA figures here: connection limits are site-specific and only a Western Power connection assessment for your actual address settles them. What we can say generically is that the further a property sits from a strong feeder, the more a battery-backed system changes the conversation, because it converts a connection-capacity problem into a design problem the installer can solve. If an off-grid or hybrid outcome is on the table, our off-grid system cost guide covers what the money actually buys and our generator-hybrid sizing guide covers keeping a genset in the design honestly.

Here is the honest part: there is no published, verified installed-price-per-kWh table for Australian winery solar-plus-storage, and anyone quoting one precise payback for your site is guessing. What the public record supports is directional. The California CEC demonstration returned about $12,000 per year on a 200 kW / 280 kWh system (CEC-500-2022-006, 2022). Korbel offset 90% of its average bill (EnergyToolBase case study, 2022). Domaine Carneros modelled roughly a ten-year return (Microgrid Knowledge). Adjacent Australian evidence from dairy case studies of 30-100 kW solar with about 200 kWh of storage quotes paybacks in the 5.9 to 7.9 year range (AU case studies, Dairy Australia(opens in a new tab); treat those as class-mates, not winery quotes.

Two levers move the answer more than panel price. First, tariff structure: a system that shaves the demand charge and fits the crush inside an existing connection cap is worth more than raw kWh offset. Second, avoided cost: Lowe Wines' off-grid decision removed a grid connection entirely (ABC, 2025), and our Harvey farm case study avoided about $200,000 in a grid-connection quote at residential-farm scale, the same mechanism at winery scale. Get a real quote against your interval data; that is the only payback that counts.

A winery-grade quote should itemise what the cheap quotes hide: the load audit it is based on, the overload and fault currents the inverter-charger will actually deliver, battery module-level expansion limits, generator start and charge integration, monitoring and telemetry, and who owns commissioning and warranty service in your region. If a quote is a single line with a kW figure, it has not been engineered against a crush.

The Lowe Wines off-grid project generated national media coverage, with owner David Lowe describing the shift from being captive to the grid as being in control (ABC, 2025). That is the pattern: energy architecture at a winery is a cellar-door story, a wedding-venue story and a sustainability-submission story, not just an engineering line item.

In the vineyard itself, agrivoltaics is the emerging adjacent story, and the research verdict is genuinely interesting rather than hyped. **Agrivoltaics** is the co-location of solar panels above or between crops so the same land produces both food and electricity. A six-year Sun'Agri/INRAE dynamic agrivoltaic trial on vines at Piolenc, France (2019-2024) found shading delayed veraison by 2-5 days and harvest by up to 7 days while retaining acidity at similar sugar levels (IVES Open Science(opens in a new tab). For a warming climate that is a potential adaptation tool, and for a winery it is a marketing story with published data behind it. It is not yet an energy product with an Australian supply chain, treat it as research-grade.